The MIT-BIH Polysomnographic Database is a collection of recordings of multiple physiologic signals during sleep. Subjects were monitored in Boston's Beth Israel Hospital Sleep Laboratory for evaluation of chronic obstructive sleep apnea syndrome, and to test the effects of constant positive airway pressure (CPAP), a standard therapeutic intervention that usually prevents or substantially reduces airway obstruction in these subjects.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":663056564,"compressed_storage_size":663076626,"resource_type":"Database","access_policy":"Open","topics":["sleep","eeg","ecg","multiparameter","respiration","apnea"],"source_url":"https://physionet.org/content/slpdb/1.0.0/"},{"public_project_uuid":"05c8b51b-9a45-4e90-bf96-6cce6d6c7773","slug":"stdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2ZW2H","is_latest_version":true,"short_description":"Twenty eight ECG recordings of varying lengths, most of which were recorded during exercise stress tests and which exhibit transient ST depression.","publish_date":"1999-08-03","title":"MIT-BIH ST Change Database","abstract":"This database includes 28 ECG recordings of varying lengths, most of which were recorded during exercise stress tests and which exhibit transient ST depression. The last five records (323 through 327) are excerpts of long-term ECG recordings and exhibit ST elevation.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":44053054,"compressed_storage_size":44075460,"resource_type":"Database","access_policy":"Open","topics":["ecg","stress","exercise"],"source_url":"https://physionet.org/content/stdb/1.0.0/"},{"public_project_uuid":"38bf8d6d-5c80-473d-87cf-926a254f5c12","slug":"cdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2QP41","is_latest_version":true,"short_description":"","publish_date":"1999-08-03","title":"MIT-BIH ECG Compression Test Database","abstract":"This database contains 168 short ECG recordings (20.48 seconds each) selected to pose a variety of challenges for ECG compressors, in particular for lossy compression methods.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":2693849,"compressed_storage_size":2821283,"resource_type":"Database","access_policy":"Open","topics":["ecg"],"source_url":"https://physionet.org/content/cdb/1.0.0/"},{"public_project_uuid":"5162a9a5-6e00-4a1f-ad04-8306d521f53b","slug":"ltdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2KS3F","is_latest_version":true,"short_description":"","publish_date":"1999-08-03","title":"MIT-BIH Long-Term ECG Database","abstract":"The MIT-BIH Long-Term ECG Database is a collection of 7 long-duration electrocardiogram (ECG) recordings (14 to 22 hours each), with manually reviewed beat annotations. The database was primarily developed to support research in the field of cardiac arrhythmia detection and analysis, and is a collaborative effort between the Massachusetts Institute of Technology (MIT) and the Beth Israel Hospital (BIH).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":215418763,"compressed_storage_size":215425191,"resource_type":"Database","access_policy":"Open","topics":["ecg"],"source_url":"https://physionet.org/content/ltdb/1.0.0/"},{"public_project_uuid":"516d30e4-cfe6-4fd3-8971-c9524887b7e3","slug":"vfdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C22P44","is_latest_version":true,"short_description":"","publish_date":"1999-08-03","title":"MIT-BIH Malignant Ventricular Ectopy Database","abstract":"This database includes 22 half-hour ECG recordings of subjects who experienced episodes of sustained ventricular tachycardia, ventricular flutter, and ventricular fibrillation.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":34683224,"compressed_storage_size":34705640,"resource_type":"Database","access_policy":"Open","topics":["ectopy","ecg"],"source_url":"https://physionet.org/content/vfdb/1.0.0/"},{"public_project_uuid":"a42b3b75-3c88-4a63-a8eb-c5f593ab0cfe","slug":"nstdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2HS3T","is_latest_version":true,"short_description":"","publish_date":"1999-08-03","title":"MIT-BIH Noise Stress Test Database","abstract":"This database includes 12 half-hour ECG recordings and 3 half-hour recordings of noise typical in ambulatory ECG recordings. The noise recordings were made using physically active volunteers and standard ECG recorders, leads, and electrodes; the electrodes were placed on the limbs in positions in which the subjects’ ECGs were not visible.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":70927689,"compressed_storage_size":70945297,"resource_type":"Database","access_policy":"Open","topics":["ecg","noise"],"source_url":"https://physionet.org/content/nstdb/1.0.0/"},{"public_project_uuid":"c8db4ca5-96d5-4f8b-b68b-d96a17fe9b64","slug":"svdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2V30W","is_latest_version":true,"short_description":"","publish_date":"1999-08-03","title":"MIT-BIH Supraventricular Arrhythmia Database","abstract":"This database includes 78 half-hour ECG recordings chosen to supplement the examples of supraventricular arrhythmias in the MIT-BIH Arrhythmia Database.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":54496449,"compressed_storage_size":54576759,"resource_type":"Database","access_policy":"Open","topics":["ecg","arrhythmia"],"source_url":"https://physionet.org/content/svdb/1.0.0/"},{"public_project_uuid":"74239889-b28f-4728-910e-915140e04be7","slug":"nsrdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2NK5R","is_latest_version":true,"short_description":"Long-term ECG recordings of 18 subjects referred to the Arrhythmia Laboratory at Boston's Beth Israel Hospital.","publish_date":"1999-08-03","title":"MIT-BIH Normal Sinus Rhythm Database","abstract":"This database includes 18 long-term ECG recordings of subjects referred to the Arrhythmia Laboratory at Boston's Beth Israel Hospital (now the Beth Israel Deaconess Medical Center). Subjects included in this database were found to have had no significant arrhythmias; they include 5 men, aged 26 to 45, and 13 women, aged 20 to 50.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":610990291,"compressed_storage_size":611007827,"resource_type":"Database","access_policy":"Open","topics":["sinus","normal","ecg"],"source_url":"https://physionet.org/content/nsrdb/1.0.0/"},{"public_project_uuid":"5fdc8b7e-ac94-4a19-ac3d-0ceef5065c0f","slug":"excluded","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2W88Z","is_latest_version":true,"short_description":"","publish_date":"1999-08-03","title":"Recordings excluded from the NSR DB","abstract":"The recordings in this collection were originally selected for inclusion in the MIT-BIH Normal Sinus Rhythm Database, but were excluded after study revealed low-grade arrhythmias.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":37513342,"compressed_storage_size":37515330,"resource_type":"Database","access_policy":"Open","topics":["ecg","arrhythmia"],"source_url":"https://physionet.org/content/excluded/1.0.0/"},{"public_project_uuid":"c3be2672-88c5-40d5-acfe-afa9eb8e29fe","slug":"szdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2QC72","is_latest_version":true,"short_description":"","publish_date":"1999-10-22","title":"Post-Ictal Heart Rate Oscillations in Partial Epilepsy","abstract":"This database contains “post-ictal heart rate oscillations in a heterogeneous group of patients with partial epilepsy.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":24326026,"compressed_storage_size":24334646,"resource_type":"Database","access_policy":"Open","topics":["ecg","epilepsy"],"source_url":"https://physionet.org/content/szdb/1.0.0/"},{"public_project_uuid":"d375020a-bfba-4575-8429-7932013ac815","slug":"gait-maturation-db","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2H014","is_latest_version":true,"short_description":"","publish_date":"1999-10-28","title":"Gait Maturation Database","abstract":"The data contained here are from 50 healthy children ranging in age from 40 months to 163 months. Each data file is named with a subject identifier (1-50) and the subject's age (e.g., the file name 2-45.txt indicates that subject #2 is 45 months old). Each data file contains two columns of data in text format. The first column is the time (in seconds), and the second column is the stride interval (in seconds) or the gait cycle duration, the time from heel strike to heel strike of the same foot.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1454243,"compressed_storage_size":1489329,"resource_type":"Database","access_policy":"Open","topics":["gait"],"source_url":"https://physionet.org/content/gait-maturation-db/1.0.0/"},{"public_project_uuid":"ceeaff97-1e28-49f2-a776-d7b1a98f1dee","slug":"qtdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C24K53","is_latest_version":true,"short_description":"One hundred two-lead ECG recordings, many extracted from other databases, with onset, peak, and end markers for P, QRS, T, and U waves.","publish_date":"1999-11-16","title":"QT Database","abstract":"Over 100 fifteen-minute two-lead ECG recordings (many excerpted from other databases), with onset, peak, and end markers for P, QRS, T, and (where present) U waves of from 30 to 50 selected beats in each recording.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":86968462,"compressed_storage_size":87119682,"resource_type":"Database","access_policy":"Open","topics":["ecg"],"source_url":"https://physionet.org/content/qtdb/1.0.0/"},{"public_project_uuid":"d006c631-86ad-460b-8b6c-3d49d857cbb6","slug":"santa-fe","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C20W2T","is_latest_version":true,"short_description":"","publish_date":"2000-01-06","title":"Santa Fe Time Series Competition Data Set B","abstract":"This is a multivariate data set recorded from a patient in the sleep laboratory of the Beth Israel Hospital (now the Beth Israel Deaconess Medical Center) in Boston, Massachusetts. This data set was extracted from record slp60 of the MIT-BIH Polysomnographic Database, and it was submitted to the Santa Fe Time Series Competition in 1991 by our group. The data are presented in text form and have been split into two sequential parts. Each line contains simultaneous samples of three parameters; the interval between samples in successive lines is 0.5 seconds. The first column is the heart rate, the second is the chest volume (respiration force), and the third is the blood oxygen concentration (measured by ear oximetry). The sampling frequency for each measurement is 2 Hz (i.e., the time interval between measurements in successive rows is 0.5 seconds).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":586703,"compressed_storage_size":587305,"resource_type":"Database","access_policy":"Open","topics":["sleep","multiparameter"],"source_url":"https://physionet.org/content/santa-fe/1.0.0/"},{"public_project_uuid":"e616e6e5-b398-408c-b08f-d39eebd36dde","slug":"apnea-ecg","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C23W2R","is_latest_version":true,"short_description":"Seventy ECG signals with expert-labelled apnea annotations and machine-generated QRS annotations.","publish_date":"2000-02-10","title":"Apnea-ECG Database","abstract":"The data consist of 70 records, divided into a learning set of 35 records (a01 through a20, b01 through b05, and c01 through c10), and a test set of 35 records (x01 through x35), all of which may be downloaded from this page. Recordings vary in length from slightly less than 7 hours to nearly 10 hours each. Each recording includes a continuous digitized ECG signal, a set of apnea annotations (derived by human experts on the basis of simultaneously recorded respiration and related signals), and a set of machine-generated QRS annotations (in which all beats regardless of type have been labeled normal). In addition, eight recordings (a01 through a04, b01, and c01 through c03) are accompanied by four additional signals (Resp C and Resp A, chest and abdominal respiratory effort signals obtained using inductance plethysmography; Resp N, oronasal airflow measured using nasal thermistors; and SpO2, oxygen saturation).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":608786550,"compressed_storage_size":608841016,"resource_type":"Database","access_policy":"Open","topics":["multiparameter","challenge","ecg","sleep","apnea"],"source_url":"https://physionet.org/content/apnea-ecg/1.0.0/"},{"public_project_uuid":"852ebcdb-5d8c-4ed0-92b9-d1ccc175e492","slug":"mimicdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2JS34","is_latest_version":true,"short_description":"","publish_date":"2000-03-15","title":"MIMIC Database","abstract":"The MIMIC Database includes data recorded from over 90 ICU patients. The data in each case include signals and periodic measurements obtained from a bedside monitor as well as clinical data obtained from the patient's medical record. The recordings vary in length; almost all of them are at least 20 hours, and many are 40 hours or more. In all, the database contains nearly 200 patient-days of real-time signals and accompanying data.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":25193606994,"compressed_storage_size":25203405584,"resource_type":"Database","access_policy":"Open","topics":["critical care","icu","health record","mimic","ehr"],"source_url":"https://physionet.org/content/mimicdb/1.0.0/"},{"public_project_uuid":"32c3a0a3-5329-4340-b706-49e409cf50e6","slug":"meditation","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2B880","is_latest_version":true,"short_description":"","publish_date":"2000-10-13","title":"Heart Rate Oscillations during Meditation","abstract":"Heart rate time series for 5 different groups of healthy subjects performing meditation techniques.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":9163842,"compressed_storage_size":9203140,"resource_type":"Database","access_policy":"Open","topics":["interbeat","heart rate","meditation","rr interval"],"source_url":"https://physionet.org/content/meditation/1.0.0/"},{"public_project_uuid":"d710ef73-89ee-4dd4-a92a-8ace4f048a0e","slug":"gaitdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2C889","is_latest_version":true,"short_description":"Walking stride interval time series from 15 subjects.","publish_date":"2000-10-13","title":"Gait in Aging and Disease Database","abstract":"Walking stride interval time series included are from 15 subjects: 5 healthy young adults (23 - 29 years old), 5 healthy old adults (71 - 77 years old), and 5 older adults (60 - 77 years old) with Parkinson's disease. The file name indicates old (o), young (y) or Parkinson's disease (pd). For the old and young subjects, the age (in years) is also included in the filename.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":363148,"compressed_storage_size":372572,"resource_type":"Database","access_policy":"Open","topics":["aging","gait"],"source_url":"https://physionet.org/content/gaitdb/1.0.0/"},{"public_project_uuid":"a8fdb442-89f4-4fb5-8c44-b3fa7574039a","slug":"chfdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C29G60","is_latest_version":true,"short_description":"Long-term ECG recordings from 15 subjects with severe congestive heart failure.","publish_date":"2000-10-14","title":"BIDMC Congestive Heart Failure Database","abstract":"This database includes long-term ECG recordings from 15 subjects (11 men, aged 22 to 71, and 4 women, aged 54 to 63) with severe congestive heart failure (NYHA class 3–4). This group of subjects was part of a larger study group receiving conventional medical therapy prior to receiving the oral inotropic agent, milrinone.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":810980216,"compressed_storage_size":608715569,"resource_type":"Database","access_policy":"Open","topics":["congestive heart failure","chf","ecg"],"source_url":"https://physionet.org/content/chfdb/1.0.0/"},{"public_project_uuid":"8801872b-1df7-4f53-b02e-e819e8467530","slug":"afdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2MW2D","is_latest_version":true,"short_description":"","publish_date":"2000-11-04","title":"MIT-BIH Atrial Fibrillation Database","abstract":"This database includes 25 long-term ECG recordings of human subjects with atrial fibrillation (mostly paroxysmal).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":635350603,"compressed_storage_size":461033254,"resource_type":"Database","access_policy":"Open","topics":["ecg","atrial fibrillation"],"source_url":"https://physionet.org/content/afdb/1.0.0/"},{"public_project_uuid":"9cfec7b3-9043-44ca-a598-22685fbbee8a","slug":"ltstdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2G01T","is_latest_version":true,"short_description":"","publish_date":"2000-11-29","title":"Long Term ST Database","abstract":"The Long-Term ST Database contains 86 lengthy ECG recordings of 80 human subjects, chosen to exhibit a variety of events of ST segment changes, including ischemic ST episodes, axis-related non-ischemic ST episodes, episodes of slow ST level drift, and episodes containing mixtures of these phenomena. The database was created to support development and evaluation of algorithms capable of accurate differentiation of ischemic and non-ischemic ST events, as well as basic research into mechanisms and dynamics of myocardial ischemia.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":10246497539,"compressed_storage_size":10246923871,"resource_type":"Database","access_policy":"Open","topics":["ecg"],"source_url":"https://physionet.org/content/ltstdb/1.0.0/"},{"public_project_uuid":"8c22634c-0be2-437f-8ed9-4fc7c965b5ea","slug":"gaitndd","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C27G6C","is_latest_version":true,"short_description":"Database of force-sensitive resistors (with the output roughly proportional to the force under the foot) from patients from Parkinson's disease (n=15), Huntington's disease (n=20), amyotrophic lateral sclerosis (n=13), and healthy subjects (n=16).","publish_date":"2000-12-21","title":"Gait in Neurodegenerative Disease Database","abstract":"Neuro-degenerative disease often affects gait and mobility. To understand better the pathophysiology of these diseases and to improve our ability to measure responses to therapeutic interventions, it may be helpful to quantify gait dynamics accurately. The records in this database are from patients with Parkinson's disease (n = 15), Huntington's disease (n = 20), or amyotrophic lateral sclerosis (n = 13). Records from 16 healthy control subjects are also included here.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":18725614,"compressed_storage_size":18788912,"resource_type":"Database","access_policy":"Open","topics":["neurodegeneration","gait"],"source_url":"https://physionet.org/content/gaitndd/1.0.0/"},{"public_project_uuid":"fed34bbb-2eed-4250-a8df-12c9dae60219","slug":"afpdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2H59W","is_latest_version":true,"short_description":"ECG recordings created for use in the Computers in Cardiology Challenge 2001, a competition with the goal of developing automated methods for predicting paroxysmal atrial fibrillation.","publish_date":"2001-03-01","title":"PAF Prediction Challenge Database","abstract":"This database of two-channel ECG recordings has been created for use in the Computers in Cardiology Challenge 2001, an open competition with the goal of developing automated methods for predicting paroxysmal atrial fibrillation (PAF). See the challenge announcement for information about the competition, and see Predicting Onset of Atrial Fibrillation for a brief overview of the clinical problem, its significance, and suggestions for further reading on the subject.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":200748253,"compressed_storage_size":200903195,"resource_type":"Database","access_policy":"Open","topics":["challenge","ecg"],"source_url":"https://physionet.org/content/afpdb/1.0.0/"},{"public_project_uuid":"66a7dc53-4281-4986-badb-a98778ccf748","slug":"umwdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C28679","is_latest_version":true,"short_description":"","publish_date":"2001-08-16","title":"Long-term Recordings of Gait Dynamics","abstract":"Stride interval fluctuations were studied in ten young, healthy men. Participants had no history of any neuromuscular, respiratory or cardiovascular disorders, and were taking no medications. Mean age was 21.7 years (range: 18-29 years). Height was 1.77 ± 0.08 meters (mean ± S.D.) and weight was 71.8 ± 10.7 kg. All subjects provided informed written consent.
\r\n\r\nSubjects walked continuously on level ground around an obstacle free, long (either 225 or 400 meters), approximately oval path and the stride interval was measured using ultra-thin, force sensitive switches taped inside one shoe.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1191758,"compressed_storage_size":1203366,"resource_type":"Database","access_policy":"Open","topics":["gait"],"source_url":"https://physionet.org/content/umwdb/1.0.0/"},{"public_project_uuid":"f0d805e7-8e52-4bf5-a948-186118210238","slug":"images","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2001G","is_latest_version":true,"short_description":"","publish_date":"2001-08-17","title":"Samples of MR Images","abstract":"These magnetic resonance angiography (MRA) images show coronal slices acquired from consecutive anteroposterior positions within the torso. The study was performed on a 1.5T General Electric (GE) Signa imaging system with gadolinium-contrast-enhancement for visualization of the cardiopulmonary vasculature. The pulse sequence used was a 3D time-of-flight fast spoiled gradient recalled acquisition in steady state (FSPGR, TR=6.3, TE=1.4, NEX=1, FOV = 40cm, slice thickness = 1.2mm).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":74234317,"compressed_storage_size":74247741,"resource_type":"Database","access_policy":"Open","topics":["image","angiography"],"source_url":"https://physionet.org/content/images/1.0.0/"},{"public_project_uuid":"ee3f651e-8c02-450c-9454-935fe27a2422","slug":"tremordb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2D097","is_latest_version":true,"short_description":"Rest tremor velocity in the index finger of 16 subjects with Parkinson's disease who receive chronic high frequency electrical deep brain stimulation.","publish_date":"2001-08-27","title":"Effect of Deep Brain Stimulation on Parkinsonian Tremor","abstract":"The recordings of this database are of rest tremor velocity in the index finger of 16 subjects with Parkinson's disease (PD) who receive chronic high frequency electrical deep brain stimulation (DBS).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":3625870,"compressed_storage_size":3654880,"resource_type":"Database","access_policy":"Open","topics":["brain","parkinsons","neuroelectric"],"source_url":"https://physionet.org/content/tremordb/1.0.0/"},{"public_project_uuid":"2a0b4eeb-a253-4e06-87de-52cc39323323","slug":"mghdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C26K5Q","is_latest_version":true,"short_description":"","publish_date":"2002-06-10","title":"MGH/MF Waveform Database","abstract":"The Massachusetts General Hospital/Marquette Foundation (MGH/MF) Waveform Database is a comprehensive collection of electronic recordings of hemodynamic and electrocardiographic waveforms of stable and unstable patients in critical care units, operating rooms, and cardiac catheterization laboratories. It is the result of a collaboration between physicians, biomedical engineers and nurses at the Massachusetts General Hospital. The database consists of recordings from 250 patients and represents a broad spectrum of physiologic and pathophysiologic states. Individual recordings vary in length from 12 to 86 minutes, and in most cases are about an hour long.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":4522848200,"compressed_storage_size":4522973710,"resource_type":"Database","access_policy":"Open","topics":["multiparameter","critical care","ecg","mgh"],"source_url":"https://physionet.org/content/mghdb/1.0.0/"},{"public_project_uuid":"683535f1-2431-4244-9d1c-d9aba39e2b2f","slug":"sleep-edf","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2C30J","is_latest_version":true,"short_description":"","publish_date":"2002-12-12","title":"Sleep-EDF Database","abstract":"The sleep-edf database contains whole-night PolySomnoGraphic sleep recordings, containing EEG, EOG, chin EMG, and event markers.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":329911165,"compressed_storage_size":329913881,"resource_type":"Database","access_policy":"Open","topics":["sleep","multiparameter","polysomnogram"],"source_url":"https://physionet.org/content/sleep-edf/1.0.0/"},{"public_project_uuid":"c02e7ca2-f336-4014-90f4-a7192800bc8d","slug":"aami-ec13","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2WC7M","is_latest_version":true,"short_description":"","publish_date":"2003-01-03","title":"ANSI/AAMI EC13 Test Waveforms","abstract":"The files in this set can be used for testing a variety of devices that monitor the electrocardiogram. The recordings include both synthetic and real waveforms.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1113137,"compressed_storage_size":1116845,"resource_type":"Database","access_policy":"Open","topics":["ecg","ansi","standard"],"source_url":"https://physionet.org/content/aami-ec13/1.0.0/"},{"public_project_uuid":"f46ce8ff-275e-486f-b636-b68b6861d2f6","slug":"fantasia","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2RG61","is_latest_version":true,"short_description":"ECG and respiration signals collected from 40 young and elderly subjects during supine resting.","publish_date":"2003-03-02","title":"Fantasia Database","abstract":"Twenty young (21 - 34 years old) and twenty elderly (68 - 85 years old) rigorously-screened healthy subjects underwent 120 minutes of continuous supine resting while continuous electrocardiographic (ECG), and respiration signals were collected; in half of each group, the recordings also include an uncalibrated continuous non-invasive blood pressure signal. Each subgroup of subjects includes equal numbers of men and women.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":306741009,"compressed_storage_size":306763953,"resource_type":"Database","access_policy":"Open","topics":["multiparameter","respiration","ecg"],"source_url":"https://physionet.org/content/fantasia/1.0.0/"},{"public_project_uuid":"e4e3f54e-b1bc-410a-8d23-10d87891cd08","slug":"chf2db","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2F598","is_latest_version":true,"short_description":"","publish_date":"2003-03-03","title":"Congestive Heart Failure RR Interval Database","abstract":"This database includes beat annotation files for 29 long-term ECG recordings of subjects aged 34 to 79, with congestive heart failure (NYHA classes I, II, and III).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":6696143,"compressed_storage_size":6708569,"resource_type":"Database","access_policy":"Open","topics":["interbeat","rr interval","chr"],"source_url":"https://physionet.org/content/chf2db/1.0.0/"},{"public_project_uuid":"77cf75d0-6d37-4468-bad8-108784c64e48","slug":"nsr2db","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2S881","is_latest_version":true,"short_description":"Beat annotation files for 54 long-term ECG recordings of subjects in normal sinus rhythm.","publish_date":"2003-03-03","title":"Normal Sinus Rhythm RR Interval Database","abstract":"This database includes beat annotation files for 54 long-term ECG recordings of subjects in normal sinus rhythm (30 men, aged 28.5 to 76, and 24 women, aged 58 to 73). The original ECG recordings (not available) were digitized at 128 samples per second, and the beat annotations were obtained by automated analysis with manual review and correction.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":11666005,"compressed_storage_size":11687119,"resource_type":"Database","access_policy":"Open","topics":["sinus","interbeat","normal","rr interval"],"source_url":"https://physionet.org/content/nsr2db/1.0.0/"},{"public_project_uuid":"d5394c31-b667-488f-b112-e462fbb4f5c9","slug":"iafdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C23S33","is_latest_version":true,"short_description":"","publish_date":"2003-03-04","title":"Intracardiac Atrial Fibrillation Database","abstract":"This database consists of endocardial recordings from the right atria of 8 patients in atrial fibrillation or flutter. A decapolar catheter with 2-5-2mm spacing (7mm spacing between bipoles) was placed in four separate regions of the heart. In each region, the 5 bipolar signals were recorded along with 3 surface ECG leads.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":107787832,"compressed_storage_size":107808500,"resource_type":"Database","access_policy":"Open","topics":["ecg","atrial fibrillation"],"source_url":"https://physionet.org/content/iafdb/1.0.0/"},{"public_project_uuid":"1547919f-838b-4055-82e5-0bcb2658aec6","slug":"tns","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2KK54","is_latest_version":true,"short_description":"Data collected for a study on a scaling analysis method used to estimate long-range power-law correlation exponents in noisy signals.","publish_date":"2003-03-07","title":"Surrogate Data with Correlations, Trends, and Nonstationarities","abstract":"The data in this collection include: (1) 6 surrogate stationary signals with different correlations; (2) 7 surrogate correlated signals with linear, sinusoidal and power-law trends; and (3) 15 surrogate correlated signals with different types of nonstationarities. Each data file contains one column of data in ASCII format.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":18948577,"compressed_storage_size":19137693,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/tns/1.0.0/"},{"public_project_uuid":"e525f080-d5c1-47d2-9b5f-b04c8627732b","slug":"ahadb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C27P4P","is_latest_version":true,"short_description":"Two ECG signals that were excluded from the 1980 American Heart Association database.","publish_date":"2003-06-06","title":"AHA Database Sample Excluded Record","abstract":"The files available here are not part of the AHA Database. The 0001.* and 0201.* files have been derived from a sample given to us in 1980 by the creators of the AHA Database. The original recording had been a candidate for inclusion in the database and was digitized and annotated using the same methods used for the AHA Database records. After the annotation process was complete, the recording was excluded from the AHA Database since it was found to contain ectopy of a higher grade than the class for which it had been chosen as a candidate.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":9686679,"compressed_storage_size":9688697,"resource_type":"Database","access_policy":"Open","topics":["ecg","american heart association"],"source_url":"https://physionet.org/content/ahadb/1.0.0/"},{"public_project_uuid":"c588d3a5-cc69-44cf-b813-af938e8694e0","slug":"shhpsgdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2MK5F","is_latest_version":true,"short_description":"Data collected for a prospective cohort study designed to investigate the relationship between sleep disordered breathing and cardiovascular disease.","publish_date":"2003-10-23","title":"Sleep Heart Health Study PSG Database","abstract":"The Sleep Heart Health Study (SHHS) is a prospective cohort study designed to investigate the relationship between sleep disordered breathing and cardiovascular disease.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":98148472,"compressed_storage_size":98151064,"resource_type":"Database","access_policy":"Open","topics":["sleep","multiparameter","polysomnogram"],"source_url":"https://physionet.org/content/shhpsgdb/1.0.0/"},{"public_project_uuid":"61705b6d-6bf6-4d79-8f54-ee20c3eee0a4","slug":"aftdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2CC7Z","is_latest_version":true,"short_description":"ECG recordings created for the Computers in Cardiology Challenge 2004, which focused on predicting spontaneous termination of atrial fibrillation.","publish_date":"2004-04-11","title":"AF Termination Challenge Database","abstract":"This database of two-channel ECG recordings has been created for use in the Computers in Cardiology Challenge 2004, an open competition with the goal of developing automated methods for predicting spontaneous termination of atrial fibrillation (AF).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":2504305,"compressed_storage_size":2552669,"resource_type":"Database","access_policy":"Open","topics":["challenge","ecg","atrial fibrillation"],"source_url":"https://physionet.org/content/aftdb/1.0.0/"},{"public_project_uuid":"d6e788db-01fe-4340-8761-d34d2f0056d1","slug":"sddb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2W306","is_latest_version":true,"short_description":"","publish_date":"2004-07-02","title":"Sudden Cardiac Death Holter Database","abstract":"PhysioNet has inaugurated a Sudden Cardiac Death Database to support research and to stimulate progress in this important area of electrophysiology. We initiate this database with 23 complete Holter recordings (originally collected by Scott Greenwald while he was at MIT), from which half-hour excerpts have been available to researchers since 1989 as the MIT-BIH Malignant Ventricular Arrhythmia Database. The database currently includes 18 patients with underlying sinus rhythm (4 with intermittent pacing), 1 who was continuously paced, and 4 with atrial fibrillation. All patients had a sustained ventricular tachyarrhythmia, and most had an actual cardiac arrest.
\r\n\r\n","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1212209318,"compressed_storage_size":1212228596,"resource_type":"Database","access_policy":"Open","topics":["holter","ecg","death"],"source_url":"https://physionet.org/content/sddb/1.0.0/"},{"public_project_uuid":"0ae2e75a-a0c4-4d98-b45c-202bf2f0f06f","slug":"crisdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C25P42","is_latest_version":true,"short_description":"Data from the Cardiac Arrhythmia Suppression Trial (CAST), a study designed to test the hypothesis that suppression of ventricular premature complexes would improve survival.","publish_date":"2004-07-02","title":"CAST RR Interval Sub-Study Database","abstract":"
The Cardiac Arrhythmia Suppression Trial (CAST) was a landmark NHLBI-sponsored study designed to test the hypothesis that the suppression of asymptomatic or mildly symptomatic ventricular premature complexes (PVCs) in survivors of myocardial infarction (MI) would decrease the number of deaths from ventricular arrhythmias and improve survival. Enrollment required an acute MI within the preceding 2 years and 6 or more PVCs per hour during a pre-treatment (qualifying) long-term ECG (Holter) recording. Those subjects enrolled within 90 days of the index MI were required to have left ventricular ejection fractions less than or equal to 55%, while those enrolled after this 90 day window were required to have an ejection fraction less than or equal to 40%. CAST enrolled 3,549 patients in all.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":645335908,"compressed_storage_size":646530296,"resource_type":"Database","access_policy":"Open","topics":["interbeat","rr interval"],"source_url":"https://physionet.org/content/crisdb/1.0.0/"},{"public_project_uuid":"75c7fde7-3bc9-4e87-bab1-bdbcf3a9e4e0","slug":"ptbdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C28C71","is_latest_version":true,"short_description":"ECGs obtained from 290 subjects using a non-commercial, prototype recorder developed at Physikalisch-Technische Bundesanstalt.","publish_date":"2004-09-25","title":"PTB Diagnostic ECG Database","abstract":"Physikalisch-Technische Bundesanstalt (PTB), the National Metrology Institute of Germany, has provided this compilation of digitized ECGs for research, algorithmic benchmarking or teaching purposes to the users of PhysioNet. The ECGs were collected from healthy volunteers and patients with different heart diseases by Professor Michael Oeff, M.D.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1790142603,"compressed_storage_size":1790456359,"resource_type":"Database","access_policy":"Open","topics":["ecg"],"source_url":"https://physionet.org/content/ptbdb/1.0.0/"},{"public_project_uuid":"4ac73f3e-7c46-4fab-8d40-9613a7e79625","slug":"nesfdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C21S3G","is_latest_version":true,"short_description":"Postural sway measurements for 27 healthy young and elderly volunteers.","publish_date":"2004-10-29","title":"Noise Enhancement of Sensorimotor Function","abstract":"This database contains postural sway measurements for 15 healthy young (mean age 23, standard deviation 2), and 12 healthy elderly (mean age 73, standard deviation 3) volunteers. Each subject's postural sway was recorded during a test of 10 minutes for the young subjects, or 5 minutes for the elderly subjects, in all cases with a 2-minute seated break midway through the test. Each test was divided into 30-second trials, and each file of the database contains data for one of these 30-second trials.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":33806076,"compressed_storage_size":33903700,"resource_type":"Database","access_policy":"Open","topics":["noise","gait"],"source_url":"https://physionet.org/content/nesfdb/1.0.0/"},{"public_project_uuid":"3420ba3c-1992-432a-b7ef-004b8f06bcef","slug":"mitdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2F305","is_latest_version":true,"short_description":"Two-channel ambulatory ECG recordings, obtained from 47 subjects studied by the BIH Arrhythmia Laboratory between 1975 and 1979.","publish_date":"2005-02-24","title":"MIT-BIH Arrhythmia Database","abstract":"The MIT-BIH Arrhythmia Database contains 48 half-hour excerpts of two-channel ambulatory ECG recordings, obtained from 47 subjects studied by the BIH Arrhythmia Laboratory between 1975 and 1979. Twenty-three recordings were chosen at random from a set of 4000 24-hour ambulatory ECG recordings collected from a mixed population of inpatients (about 60%) and outpatients (about 40%) at Boston's Beth Israel Hospital; the remaining 25 recordings were selected from the same set to include less common but clinically significant arrhythmias that would not be well-represented in a small random sample.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":109337751,"compressed_storage_size":77030320,"resource_type":"Database","access_policy":"Open","topics":["ecg","arrhythmia"],"source_url":"https://physionet.org/content/mitdb/1.0.0/"},{"public_project_uuid":"29b4fd12-0e33-4ffd-9f05-2d6054bc5b05","slug":"cudb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2X59M","is_latest_version":true,"short_description":"","publish_date":"2007-05-02","title":"CU Ventricular Tachyarrhythmia Database","abstract":"This database includes 35 eight-minute ECG recordings of human subjects who experienced episodes of sustained ventricular tachycardia, ventricular flutter, and ventricular fibrillation.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":6749077,"compressed_storage_size":6783051,"resource_type":"Database","access_policy":"Open","topics":["ventricular fibrillation","ecg"],"source_url":"https://physionet.org/content/cudb/1.0.0/"},{"public_project_uuid":"7494521e-196f-434c-b7d6-49c7855e8177","slug":"mvtdb","version":"1.0","core_doi":null,"version_doi":"10.13026/C25K5D","is_latest_version":true,"short_description":"RR interval time series, recorded by implanted cardioverter defibrillators in 78 subjects.","publish_date":"2007-05-02","title":"Spontaneous Ventricular Tachyarrhythmia Database","abstract":"This database contains 135 pairs of RR interval time series, recorded by implanted cardioverter defibrillators (Medtronic Jewel PlusTM ICD 7218) in 78 subjects. Each series contains between 986 and 1022 RR intervals. One series of each pair includes a spontaneous episode of ventricular tachycardia (VT) or ventricular fibrillation (VF), and the other is a sample of the intrinsic (usually sinus) rhythm.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":2591031,"compressed_storage_size":2768367,"resource_type":"Database","access_policy":"Open","topics":["interbeat","rr interval","arrhythmia"],"source_url":"https://physionet.org/content/mvtdb/1.0/"},{"public_project_uuid":"c150242c-d32e-4f7f-8f14-c12e18e1b9c7","slug":"nifecgdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2X30H","is_latest_version":true,"short_description":"Fifty-five multichannel abdominal non-invasive fetal electrocardiogram recordings, taken from a single subject between 21 to 40 weeks of pregnancy.","publish_date":"2007-09-06","title":"Non-Invasive Fetal ECG Database","abstract":"This database contains a series of 55 multichannel abdominal non-invasive fetal electrocardiogram (FECG) recordings, taken from a single subject between 21 to 40 weeks of pregnancy. The records have variable durations, and were taken weekly (two or more records were acquired during some weeks). These records may be very useful for testing signal separation algorithms.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":390714029,"compressed_storage_size":390734369,"resource_type":"Database","access_policy":"Open","topics":["fetal","ecg"],"source_url":"https://physionet.org/content/nifecgdb/1.0.0/"},{"public_project_uuid":"c963e654-30b1-411a-973b-df38b2e38c28","slug":"ucddb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C26C7D","is_latest_version":true,"short_description":"Overnight polysomnograms with simultaneous three-channel Holter ECG, from 25 adult subjects with suspected sleep-disordered breathing.","publish_date":"2007-10-09","title":"St. Vincent's University Hospital / University College Dublin Sleep Apnea Database","abstract":"This database contains 25 full overnight polysomnograms with simultaneous three-channel Holter ECG, from adult subjects with suspected sleep-disordered breathing. A revised version of this database was posted on 1 September 2011.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1368695346,"compressed_storage_size":794211974,"resource_type":"Database","access_policy":"Open","topics":["sleep","ecg","holter","multiparameter","polysomnogram"],"source_url":"https://physionet.org/content/ucddb/1.0.0/"},{"public_project_uuid":"00066c1b-1363-4171-a0ec-2a575fc30b1c","slug":"gaitpdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C24H3N","is_latest_version":true,"short_description":"","publish_date":"2008-02-25","title":"Gait in Parkinson's Disease","abstract":"This database contains measures of gait from 93 patients with idiopathic PD (mean age: 66.3 years; 63% men), and 73 healthy controls (mean age: 66.3 years; 55% men). The database includes the vertical ground reaction force records of subjects as they walked at their usual, self-selected pace for approximately 2 minutes on level ground.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":302384480,"compressed_storage_size":302436928,"resource_type":"Database","access_policy":"Open","topics":["parkinsons","gait"],"source_url":"https://physionet.org/content/gaitpdb/1.0.0/"},{"public_project_uuid":"866d5dd6-1871-4a9d-a404-2c3b81841773","slug":"incartdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2V88N","is_latest_version":true,"short_description":"Annotated ECG recordings extracted from 32 Holter records.","publish_date":"2008-05-01","title":"St Petersburg INCART 12-lead Arrhythmia Database","abstract":"This database consists of 75 annotated recordings extracted from 32 Holter records. Each record is 30 minutes long and contains 12 standard leads, each sampled at 257 Hz, with gains varying from 250 to 1100 analog-to-digital converter units per millivolt.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":833098896,"compressed_storage_size":590842939,"resource_type":"Database","access_policy":"Open","topics":["arrhythmia","ecg","holter"],"source_url":"https://physionet.org/content/incartdb/1.0.0/"},{"public_project_uuid":"895787b6-17df-4d69-a435-83a08fd7b66b","slug":"drivedb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2SG6B","is_latest_version":true,"short_description":"","publish_date":"2008-05-15","title":"Stress Recognition in Automobile Drivers","abstract":"This database, contributed to PhysioNet by its creator, Jennifer Healey, contains a collection of multiparameter recordings from healthy volunteers, taken while they were driving on a prescribed route including city streets and highways in and around Boston, Massachusetts. The objective of the study for which these data were collected was to investigate the feasibility of automated recognition of stress on the basis of the recorded signals, which include ECG, EMG (right trapezius), GSR (galvanic skin resistance) measured on the hand and foot, and respiration.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":113930207,"compressed_storage_size":113937531,"resource_type":"Database","access_policy":"Open","topics":["multiparameter","stress","respiration","ecg"],"source_url":"https://physionet.org/content/drivedb/1.0.0/"},{"public_project_uuid":"ebc630a0-3b5f-401c-baf8-b2a950b6beba","slug":"twadb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2B59B","is_latest_version":true,"short_description":"Multichannel ECG records collected for the 2008 Computers In Cardiology Challenge.","publish_date":"2008-09-22","title":"T-Wave Alternans Challenge Database","abstract":"This database has been assembled for the PhysioNet/Computers in Cardiology Challenge 2008. It contains 100 multichannel ECG records sampled at 500 Hz with 16 bit resolution over a ± 32 mV range. The subjects include patients with myocardial infarctions, transient ischemia, ventricular tachyarrhythmias, and other risk factors for sudden cardiac death, as well as healthy controls and synthetic cases with calibrated amounts of T-wave alternans.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":110087840,"compressed_storage_size":110141804,"resource_type":"Database","access_policy":"Open","topics":["challenge","ecg","arrhythmia"],"source_url":"https://physionet.org/content/twadb/1.0.0/"},{"public_project_uuid":"b60ae1e3-12ce-4d17-9ea9-68d737df8248","slug":"ltafdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2QG6Q","is_latest_version":true,"short_description":"","publish_date":"2008-10-30","title":"Long Term AF Database","abstract":"This database includes 84 long-term ECG recordings of subjects with paroxysmal or sustained atrial fibrillation (AF). Each record contains two simultaneously recorded ECG signals digitized at 128 Hz with 12-bit resolution over a 20 mV range; record durations vary but are typically 24 to 25 hours.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":3649578582,"compressed_storage_size":1776262655,"resource_type":"Database","access_policy":"Open","topics":["ecg","atrial fibrillation"],"source_url":"https://physionet.org/content/ltafdb/1.0.0/"},{"public_project_uuid":"6794345e-5dad-4ca4-8762-63d8ab5d8580","slug":"emgdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C24S3D","is_latest_version":true,"short_description":"","publish_date":"2009-09-05","title":"Examples of Electromyograms","abstract":"An electromyogram (EMG) is a common clinical test used to assess function of muscles and the nerves that control them. EMG studies are used to help in the diagnosis and management of disorders such as the muscular dystrophies and neuropathies. Nerve conduction studies that measure how well and how fast the nerves conduct impulses are often performed in conjunction with EMG studies.
\r\n\r\nExamples of EMG studies (courtesy of Seward Rutkove, MD, Department of Neurology, Beth Israel Deaconess Medical Center/Harvard Medical School) are given here.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":5307170,"compressed_storage_size":5309270,"resource_type":"Database","access_policy":"Open","topics":["neuropathy","electromyogram"],"source_url":"https://physionet.org/content/emgdb/1.0.0/"},{"public_project_uuid":"85ab6168-0602-45fa-87fb-a863c4b86493","slug":"eegmmidb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C28G6P","is_latest_version":true,"short_description":"EEG recordings obtained from 109 volunteers.","publish_date":"2009-09-09","title":"EEG Motor Movement/Imagery Dataset","abstract":"The EEG Motor Movement/Imagery Dataset provides 64-channel electroencephalography recordings collected with the BCI2000 system during real and imagined motor tasks. Participants completed 14 experimental runs, including two one-minute baseline recordings with eyes open and eyes closed, followed by three repetitions of four two-minute task conditions. These conditions involved either executing or imagining unilateral fist movements in response to left/right visual targets, or executing or imagining bilateral fist or foot movements in response to top/bottom visual targets. EEG signals were recorded according to the international 10-10 electrode placement system at a sampling rate of 160 Hz and are provided in EDF+ format with accompanying annotation channels. Event labels identify rest periods and task onsets using three codes: T0 for rest, T1 for left-fist or both-fists movement/imagery depending on the run type, and T2 for right-fist or both-feet movement/imagery. The dataset supports research in brain-computer interfaces, motor imagery classification, movement-related EEG dynamics, and the development of signal-processing and machine-learning methods for neural decoding.","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":3614967988,"compressed_storage_size":1997435063,"resource_type":"Database","access_policy":"Open","topics":["eeg","movement","neuroelectric"],"source_url":"https://physionet.org/content/eegmmidb/1.0.0/"},{"public_project_uuid":"74c7f0ac-23a5-4c1b-809a-ac0ee18d7ee9","slug":"edb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2D59Z","is_latest_version":true,"short_description":"Annotated excerpts of ambulatory ECG recordings from 79 subjects, designed for evaluation of algorithms for analysis of ST and T-wave changes.","publish_date":"2009-11-27","title":"European ST-T Database","abstract":"The European ST-T Database is intended to be used for evaluation of algorithms for analysis of ST and T-wave changes. This database consists of 90 annotated excerpts of ambulatory ECG recordings from 79 subjects. The subjects were 70 men aged 30 to 84, and 8 women aged 55 to 71.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":487712650,"compressed_storage_size":487754650,"resource_type":"Database","access_policy":"Open","topics":["ecg"],"source_url":"https://physionet.org/content/edb/1.0.0/"},{"public_project_uuid":"13dc7098-9786-42a2-b6b4-eb8fe21d9f75","slug":"chbmit","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2K01R","is_latest_version":true,"short_description":"EEG recordings from pediatric subjects with intractable seizures, collected at the Children’s Hospital Boston.","publish_date":"2010-06-09","title":"CHB-MIT Scalp EEG Database","abstract":"This database, collected at the Children’s Hospital Boston, consists of EEG recordings from pediatric subjects with intractable seizures. Subjects were monitored for up to several days following withdrawal of anti-seizure medication in order to characterize their seizures and assess their candidacy for surgical intervention. The recordings are grouped into 23 cases and were collected from 22 subjects (5 males, ages 3–22; and 17 females, ages 1.5–19).","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":45759184625,"compressed_storage_size":45759349673,"resource_type":"Database","access_policy":"Open","topics":["medication","eeg","neuroelectric","seizure"],"source_url":"https://physionet.org/content/chbmit/1.0.0/"},{"public_project_uuid":"8530039e-f1dd-44ad-8505-0079b6da985f","slug":"earndb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2NP4D","is_latest_version":true,"short_description":"Auditory Brainstem Response and Otoacoustic Emission recordings generated as part of a study examining evoked potentials and loudness growth.","publish_date":"2011-02-03","title":"Evoked Auditory Responses in Normals","abstract":"This database, created and contributed by Michael J. Epstein and Ikaro Silva, was generated as part of a study examining evoked potentials and loudness growth. The database consists of two sets of physiological signals: auditory brainstem response (ABR) and otoacoustic emissions (OAE) and two sets of psychoacoustical estimates of loudness as a function of peak sound pressure level (peSPL).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":61062975623,"compressed_storage_size":61066658061,"resource_type":"Database","access_policy":"Open","topics":["loudness","auditory","neuroelectric"],"source_url":"https://physionet.org/content/earndb/1.0.0/"},{"public_project_uuid":"49c0d835-267b-4a3f-8510-89efb0739cd4","slug":"bpssrat","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C20597","is_latest_version":true,"short_description":"","publish_date":"2012-06-26","title":"Blood Pressure in Salt-Sensitive Dahl Rats","abstract":"This database contains continuous blood pressure recordings for 9 Dahl SS rats and 6 Dahl SS.13BN rats, under high and low salt conditions.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":3593170,"compressed_storage_size":3606198,"resource_type":"Database","access_policy":"Open","topics":["blood pressure","hypertension","cardiovascular"],"source_url":"https://physionet.org/content/bpssrat/1.0.0/"},{"public_project_uuid":"0edd7c14-ac3b-493d-82f8-2b78d7d8e969","slug":"capslpdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2VC79","is_latest_version":true,"short_description":"","publish_date":"2012-07-26","title":"CAP Sleep Database","abstract":"The CAP Sleep Database is a collection of 108 polysomnographic recordings registered at the Sleep Disorders Center of the Ospedale Maggiore of Parma, Italy. The waveforms (contained in the .edf files of the database) include at least 3 EEG channels (F3 or F4, C3 or C4 and O1 or O2, referred to A1 or A2), EOG (2 channels), EMG of the submentalis muscle, bilateral anterior tibial EMG, respiration signals (airflow, abdominal and thoracic effort and SaO2) and EKG. Additional traces include EEG bipolar traces, according to the 10-20 international system (Fp1-F3, F3-C3, C3-P3, P3-O1 and/or Fp2-F4, F4-C4, C4-P4, P4-O2).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":43071869200,"compressed_storage_size":43071921140,"resource_type":"Database","access_policy":"Open","topics":["sleep","eeg","ecg","multiparameter","apnea"],"source_url":"https://physionet.org/content/capslpdb/1.0.0/"},{"public_project_uuid":"870d12ed-ba67-4d07-81f8-6c77f26a9f0f","slug":"ob1db","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2D30V","is_latest_version":true,"short_description":"","publish_date":"2012-08-03","title":"OB-1 Fetal ECG Database","abstract":"This project is developing a set of recordings of fetal scalp electrograms and uterine muscular activity, with beat-by-beat annotations of the fetal ECG, to support studies of fetal heart rate variability.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":41216365,"compressed_storage_size":41217325,"resource_type":"Database","access_policy":"Open","topics":["fetal","ecg"],"source_url":"https://physionet.org/content/ob1db/1.0.0/"},{"public_project_uuid":"6f21ad8b-cd99-47fb-a146-f568935a72bd","slug":"adfecgdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2RP4B","is_latest_version":true,"short_description":"Multichannel fetal electrocardiogram recordings obtained from 5 different women in labor, between 38 and 41 weeks of gestation.","publish_date":"2012-08-09","title":"Abdominal and Direct Fetal ECG Database","abstract":"The research material included in the Abdominal and Direct Fetal Electrocardiogram Database contains multichannel fetal electrocardiogram (FECG) recordings obtained from 5 different women in labor, between 38 and 41 weeks of gestation. The recordings were acquired in the Department of Obstetrics at the Medical University of Silesia, by means of the KOMPOREL system for acquisition and analysis of fetal electrocardiogram (ITAM Institute, Zabrze, Poland). Each recording comprises four differential signals acquired from maternal abdomen and the reference direct fetal electrocardiogram registered from the fetal head.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":15316525,"compressed_storage_size":15318973,"resource_type":"Database","access_policy":"Open","topics":["fetal","ecg"],"source_url":"https://physionet.org/content/adfecgdb/1.0.0/"},{"public_project_uuid":"af5f7757-2a58-4d62-9576-2b7a6d8612ec","slug":"tpehgdb","version":"1.0.1","core_doi":null,"version_doi":"10.13026/C2FW2V","is_latest_version":true,"short_description":"Electrohysterogram records during regular check-ups at the University Medical Centre Ljubljana between 1997 and 2005.","publish_date":"2012-08-21","title":"Term-Preterm EHG Database","abstract":"The Electrohysterogram records (uterine EMG records) included in the Term-Preterm ElectroHysteroGram DataBase (TPEHG DB) were obtained from 1997 to 2005 at the University Medical Centre Ljubljana, Department of Obstetrics and Gynecology. The records were obtained during regular check-ups either around the 22nd week of gestation or around the 32nd week of gestation.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":257361301,"compressed_storage_size":257471447,"resource_type":"Database","access_policy":"Open","topics":["obstetrics","neuroelectric","gestation","electrohysterogram"],"source_url":"https://physionet.org/content/tpehgdb/1.0.1/"},{"public_project_uuid":"b5e8e0c9-32b3-40ee-beb5-3e5c2214b5e9","slug":"sleep-edfx","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2X676","is_latest_version":true,"short_description":"","publish_date":"2013-10-24","title":"Sleep-EDF Database Expanded","abstract":"The sleep-edf database contains 197 whole-night PolySomnoGraphic sleep recordings, containing EEG, EOG, chin EMG, and event markers. Some records also contain respiration and body temperature. Corresponding hypnograms (sleep patterns) were manually scored by well-trained technicians according to the Rechtschaffen and Kales manual, and are also available.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":8715229580,"compressed_storage_size":8715317980,"resource_type":"Database","access_policy":"Open","topics":["sleep","multiparameter","polysomnogram"],"source_url":"https://physionet.org/content/sleep-edfx/1.0.0/"},{"public_project_uuid":"0317321b-c715-4408-8b42-72064f073c12","slug":"simfpcgdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2GW25","is_latest_version":true,"short_description":"","publish_date":"2014-01-21","title":"Simulated Fetal Phonocardiograms","abstract":"This data set is a series of synthetic fetal phonocardiographic signals (PCGs) relative to different fetal states and recording conditions.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":35384185,"compressed_storage_size":35400653,"resource_type":"Database","access_policy":"Open","topics":["multiparameter","fetal","simulation","phonocardiograpy"],"source_url":"https://physionet.org/content/simfpcgdb/1.0.0/"},{"public_project_uuid":"e47b5c7f-7350-4205-9608-c10e8df5d397","slug":"ctu-uhb-ctgdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C22013","is_latest_version":true,"short_description":"552 cardiotocography records collected between 2010 and 2012 at the Czech Technical University and University Hospital in Brno.","publish_date":"2014-02-18","title":"CTU-CHB Intrapartum Cardiotocography Database","abstract":"This database, from the Czech Technical University (CTU) in Prague and the University Hospital in Brno (UHB), contains 552 cardiotocography (CTG) recordings, which were carefully selected from 9164 recordings collected between 2010 and 2012 at UHB.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":39897278,"compressed_storage_size":40114482,"resource_type":"Database","access_policy":"Open","topics":["multiparameter","fetal","heart rate","cardiotocography"],"source_url":"https://physionet.org/content/ctu-uhb-ctgdb/1.0.0/"},{"public_project_uuid":"1105f231-2051-4d95-b913-c139568634b0","slug":"motion-artifact","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2988P","is_latest_version":true,"short_description":"Examples of functional near-infrared spectroscopy and electroencephalogram recordings that have been created for evaluating artifact removal methods.","publish_date":"2014-03-03","title":"Motion Artifact Contaminated fNIRS and EEG Data","abstract":"This data collection, contributed to PhysioBank by Kevin Sweeney and colleagues at the National University of Ireland in Maynooth, contains examples of functional near-infrared spectroscopy (fNIRS) and electroencephalogram (EEG) recordings that have been created for evaluating artifact removal methods.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":681423884,"compressed_storage_size":681444776,"resource_type":"Database","access_policy":"Open","topics":["eeg","artifact","multiparameter","spectroscopy","noise"],"source_url":"https://physionet.org/content/motion-artifact/1.0.0/"},{"public_project_uuid":"f8266a58-13da-43cd-a404-2e56071b436d","slug":"erpbci","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2101S","is_latest_version":true,"short_description":"Data generated as part of a study aimed at identifying the factors limiting the performance of brain-computer interfaces based on event-related potentials.","publish_date":"2014-03-04","title":"ERP-based Brain-Computer Interface recordings","abstract":"This dataset — created and contributed by Luca Citi, Riccardo Poli, and Caterina Cinel — was generated as part of a study aimed at identifying the factors limiting the performance of brain-computer interfaces (BCIs) based on event-related potentials (ERPs), in order to improve the transfer rate and the usability of these interfaces.
\r\n\r\nIn this study, each participant was asked to spell a total of 20 characters using a traditional matrix speller (shown at left). Each record in the dataset contains the signals, triggers and annotations corresponding to a single run, i.e. the spelling of a single character. Target characters were randomly chosen before the beginning of the run. Each row and column of the standard 6x6 matrix of characters was randomly intensified without replacement for 100 ms with a gap of 50 ms, leading to a stimulus onset asynchrony (SOA) of 150 ms. During a run, each row and column flashed approximately 20 times. During that period, subjects were asked to focus on the target character and to mentally count the number of times it was highlighted. From subject 03 onwards, subjects were asked to report their final count at the end of the run.
\r\n\r\n","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":2356595125,"compressed_storage_size":1794770291,"resource_type":"Database","access_policy":"Open","topics":["brain-computer interface","neuroelectric"],"source_url":"https://physionet.org/content/erpbci/1.0.0/"},{"public_project_uuid":"2ae41f06-8d98-4bf3-b7c0-2495d2e10b48","slug":"ecgiddb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2J01F","is_latest_version":true,"short_description":"ECG recordings from 90 volunteers.","publish_date":"2014-03-06","title":"ECG-ID Database","abstract":"
The database contains 310 ECG recordings, obtained from 90 persons. Each recording contains:
\r\n\r\nThis database contains the combined measurement of ECG, breathing and seismocardiogram from 20 healthy volunteers.
\r\n\r\nThe purpose of the database was twofold:
\r\n\r\nThis database consists of 122 16-electrode EHG recordings performed on 45 pregnant women. The recordings were performed between 2008 and 2010 in Iceland.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":3928948476,"compressed_storage_size":2134623019,"resource_type":"Database","access_policy":"Open","topics":["neuroelectric","pregnancy","electrohysterogram"],"source_url":"https://physionet.org/content/ehgdb/1.0.0/"},{"public_project_uuid":"2310940d-1f65-4794-98e5-09cef31ce66f","slug":"shareedb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2RC7C","is_latest_version":true,"short_description":"Holter recordings of 139 hypertensive patients recruited at the Centre of Hypertension of the University Hospital of Naples Federico II.","publish_date":"2015-05-19","title":"Smart Health for Assessing the Risk of Events via ECG Database","abstract":"The SHAREE database was developed in order to investigate the possibility to identify hypertensive subjects at higher risk to develop vascular events based on Heart Rate Variability analysis. The research material in the SHAREE database included nominal 24-h electrocardiographic (ECG) Holter recordings of 139 hypertensive patients recruited at the Centre of Hypertension of the University Hospital of Naples Federico II, Naples, Italy.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":4447973254,"compressed_storage_size":1870380494,"resource_type":"Database","access_policy":"Open","topics":["hypertension","ecg","holter","risk","hrv"],"source_url":"https://physionet.org/content/shareedb/1.0.0/"},{"public_project_uuid":"78d262a3-dbf2-4ad2-87fc-89bf051c68d3","slug":"fpcgdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2PD5B","is_latest_version":true,"short_description":"","publish_date":"2015-12-16","title":"Fetal PCG Database","abstract":"The project collects a series of 26 fetal phonocardiographic (PCG) signals from different pregnant women during the last months of their singleton physiological pregnancies (gestational week between 31 and 40).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":20851572,"compressed_storage_size":20860832,"resource_type":"Database","access_policy":"Open","topics":["fetal","synthetic","phonocardiography","pregnancy"],"source_url":"https://physionet.org/content/fpcgdb/1.0.0/"},{"public_project_uuid":"089bf387-0de8-48a4-b908-5f4e006b35d8","slug":"macecgdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2JP4G","is_latest_version":true,"short_description":"","publish_date":"2015-12-18","title":"Motion Artifact Contaminated ECG Database","abstract":"Short duration ECG signals are recorded from a healthy 25-year-old male performing different physical activities to study the effect of motion artifacts on ECG signals and their sparsity.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1061119,"compressed_storage_size":1072957,"resource_type":"Database","access_policy":"Open","topics":["ecg","noise","artifact"],"source_url":"https://physionet.org/content/macecgdb/1.0.0/"},{"public_project_uuid":"b7d5a869-38ef-41e6-96f9-4ae788ff2ad0","slug":"prcp","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2PC7R","is_latest_version":true,"short_description":"","publish_date":"2016-02-18","title":"Physiologic Response to Changes in Posture","abstract":"A collection of physiological signals in ten healthy subjects in response to a slow tilt, a fast tilt, and a standing-up maneuver.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":56722178,"compressed_storage_size":56732424,"resource_type":"Database","access_policy":"Open","topics":["blood pressure","movement","ecg","multiparameter","posture"],"source_url":"https://physionet.org/content/prcp/1.0.0/"},{"public_project_uuid":"ee2a95b6-7e05-494d-8686-dbdf87152da8","slug":"mmgdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2ZW95","is_latest_version":true,"short_description":"Uterine magnetomyographic signals from 25 subjects recorded using a 151 channel Reproductive Assessment system.","publish_date":"2016-03-04","title":"MMG Database","abstract":"This database contains uterine magnetomyographic (MMG) signals recorded using the 151 channel SARA (SQUID Array for Reproductive Assessment) system installed at UAMS, Little Rock, USA. Recordings are taken from 25 subjects who presented themselves in the Triage unit of Labor and Delivery and were undergoing monitoring and evaluation of labor. All subjects were asked to lean forward and sit comfortably with the sensor array covering their pregnant abdomens. The duration of a recording was typically around 20 minutes.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":226191457,"compressed_storage_size":226199325,"resource_type":"Database","access_policy":"Open","topics":["magnetomyography","neuroelectric","pregnancy"],"source_url":"https://physionet.org/content/mmgdb/1.0.0/"},{"public_project_uuid":"17722af4-4753-44a3-b721-63c92c9529e0","slug":"sgamp","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C25C73","is_latest_version":true,"short_description":"","publish_date":"2016-04-15","title":"Squid Giant Axon Membrane Potential","abstract":"The SGAMP database contains single-unit neuronal recordings of North Atlantic squid (Loligo pealei) giant axons in response to stimulus currents. The membrane potential and stimulus current are given for a total of 170 trials across 8 different axons.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":945463796,"compressed_storage_size":945527792,"resource_type":"Database","access_policy":"Open","topics":["neuron","neuroelectric"],"source_url":"https://physionet.org/content/sgamp/1.0.0/"},{"public_project_uuid":"6f91100b-440d-4983-a091-769bf7759f0f","slug":"fecgsyndb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C21P4T","is_latest_version":true,"short_description":"","publish_date":"2016-04-29","title":"Fetal ECG Synthetic Database","abstract":"The FECGSYNDB is a large database of simulated adult and non-invasive fetal ECG (NI-FECG) signals, which provides a robust resource that enables reproducible research in the field. The data is generated using the FECGSYN simulator (visit website).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":35732007272,"compressed_storage_size":35736319968,"resource_type":"Database","access_policy":"Open","topics":["fetal","synthetic","ecg"],"source_url":"https://physionet.org/content/fecgsyndb/1.0.0/"},{"public_project_uuid":"babdd034-4cee-487b-8afb-4ab75ce3ba6e","slug":"hbedb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2WW2W","is_latest_version":true,"short_description":"Force platform recordings from 163 subjects undergoing stabilography tests.","publish_date":"2016-05-19","title":"Human Balance Evaluation Database","abstract":"The HBEDB contains force platform recordings from subjects undergoing stabilography tests. The subjects performed standing tasks under four different conditions: with their eyes opened or closed, while standing on a rigid or unstable surface. Each condition was tested three times, with the order of the conditions being randomized among subjects. A total of 1930 trials performed by 163 different subjects are given in this database.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":373478564,"compressed_storage_size":374174098,"resource_type":"Database","access_policy":"Open","topics":["stability","gait"],"source_url":"https://physionet.org/content/hbedb/1.0.0/"},{"public_project_uuid":"a2af673f-9c21-4936-8533-6c9bbf10da81","slug":"ltmm","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2S59C","is_latest_version":true,"short_description":"","publish_date":"2016-06-20","title":"Long Term Movement Monitoring Database","abstract":"The LTMM database contains 3-day 3D accelerometer recordings of 71 elder community residents, used to study gait, stability, and fall risk.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":22285420607,"compressed_storage_size":22285482421,"resource_type":"Database","access_policy":"Open","topics":["accelerometer","risk","stability","gait"],"source_url":"https://physionet.org/content/ltmm/1.0.0/"},{"public_project_uuid":"e4b58ae7-9a89-4121-987e-849a2be2f126","slug":"ecgrdvq","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2HP45","is_latest_version":true,"short_description":"ECGs of 22 subjects for a study aimed at comparing the effects of QT prolonging drugs versus placebo on electrophysiological parameters.","publish_date":"2016-07-26","title":"ECG Effects of Ranolazine, Dofetilide, Verapamil, and Quinidine","abstract":"The ECGRDVQ database contains multi-channel ECG recordings of 22 healthy subjects partaking in a randomized, double‐blind, 5‐period crossover clinical trial aimed at comparing the effects of four known QT prolonging drugs versus placebo on electrophysiological and other clinical parameters.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1490896779,"compressed_storage_size":1499072353,"resource_type":"Database","access_policy":"Open","topics":["ecg","medication"],"source_url":"https://physionet.org/content/ecgrdvq/1.0.0/"},{"public_project_uuid":"301b817d-da88-4136-b1de-c19b09dd198d","slug":"mssvepdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C29591","is_latest_version":true,"short_description":"","publish_date":"2016-09-03","title":"MAMEM SSVEP Database","abstract":"Released by the Information Technologies Institute (CERTH-ITI) and powered by MAMEM HORIZON 2020, the MSSVEP database contains EEG recordings of 11 subjects under the stimulation of flickering lights, used to study the steady state visually evoked potentials. The data is divided into three sets in which the experimental procedures and capturing devices differ.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":5526174291,"compressed_storage_size":5526321069,"resource_type":"Database","access_policy":"Open","topics":["eeg","neuroelectric","stimulation"],"source_url":"https://physionet.org/content/mssvepdb/1.0.0/"},{"public_project_uuid":"6b1956b0-4158-41d1-9467-fa688cfd5580","slug":"ecgdmmld","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2D016","is_latest_version":true,"short_description":"ECG from 22 subjects for a study on response of hERG potassium channel blocking drugs with and without the addition of late sodium or calcium channel blocking drugs.","publish_date":"2016-09-15","title":"ECG Effects of Dofetilide, Moxifloxacin, Dofetilide+Mexiletine, Dofetilide+Lidocaine and Moxifloxacin+Diltiazem","abstract":"The ECGDMMLD database contains data from a randomized, double-blind, 5-period crossover clinical trial in 22 healthy male and female subjects, 18 to 35 years of age, to compare the electrophysiological response of hERG potassium channel blocking drugs with and without the addition of late sodium or calcium channel blocking drugs. The 5 treatment periods are 1) dofetilide alone, 2) mexiletine with and without dofetilide, 3) lidocaine with and without dofetilide, 4) moxifloxacin with and without diltiazem and 5) placebo. During each treatment period, 12 blood samples for pharmacokinetic measurements are obtained with matched 12-lead ECG recordings.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1186773096,"compressed_storage_size":1195255758,"resource_type":"Database","access_policy":"Open","topics":["medication","ecg"],"source_url":"https://physionet.org/content/ecgdmmld/1.0.0/"},{"public_project_uuid":"6f262fb5-a45d-45c9-bd15-468f663218cb","slug":"nqmitcsxpd","version":"1.0.0","core_doi":"10.13026/g0bd-1m78","version_doi":"10.13026/C2859Q","is_latest_version":true,"short_description":"Keystroke logs collected from 85 subjects with and without Parkinson's disease.","publish_date":"2016-12-20","title":"neuroQWERTY MIT-CSXPD Dataset","abstract":"The neuroQWERTY MIT-CSXPD database contains keystroke logs collected from 85 subjects with and without parkinsons disease (PD). This dataset has been collected and analyzed in order to indicate that the routine interaction with computer keyboards can be used to detect motor signs in the early stages of PD.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":7643543,"compressed_storage_size":7673805,"resource_type":"Database","access_policy":"Open","topics":["brain","parkinsons","neuroelectric"],"source_url":"https://physionet.org/content/nqmitcsxpd/1.0.0/"},{"public_project_uuid":"79312db3-d5d9-4473-869a-996bb42433c7","slug":"charisdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C24G6F","is_latest_version":true,"short_description":"Multi-channel recordings of ECG, arterial blood pressure (ABP), and intracranial pressure (ICP) of patients diagnosed with traumatic brain injury.","publish_date":"2017-01-19","title":"CHARIS database","abstract":"The CHARIS database contains multi-channel recordings of ECG, arterial blood pressure (ABP), and intracranial pressure (ICP) of patients diagnosed with traumatic brain injury (TBI). The data is contributed by members of the CHARIS project which aims to systematize the analysis of relevant physiological signals, and create data-driven algorithms to search for potential predictors of acute clinical events for patients with acute brain injury.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1762910168,"compressed_storage_size":1762914180,"resource_type":"Database","access_policy":"Open","topics":["multiparameter","ecg","intracranial pressure","blood pressure"],"source_url":"https://physionet.org/content/charisdb/1.0.0/"},{"public_project_uuid":"f90ef629-deaa-48b5-9fdf-bcfb7c48c5a6","slug":"staffiii","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C20P4H","is_latest_version":true,"short_description":"","publish_date":"2017-01-31","title":"STAFF III Database","abstract":"The STAFF III database was acquired during 1995–96 at Charleston Area Medical Center (WV, USA) where single prolonged balloon inflation had been introduced to achieve optimal results of percutaneous transluminal coronary angiography (PTCA) procedures, replacing the typical series of brief inflations. The lead investigator Dr. Stafford Warren designed the study protocol together with Dr. Galen Wagner at Duke University Medical Center (Durham, NC, USA); Dr. Michael Ringborn (Blekinge Hospital, Karlskrona, Sweden) was responsible for data acquisition. The database consists of ECG recordings from 104 patients, accounting for substantial inter-patient variability in reaction to prolonged balloon inflation as well as variability of heart rhythm and waveform morphology. Only patients receiving elective PTCA in one of the major coronary arteries were included. Patients suffering from ventricular tachycardia, undergoing an emergency procedure, or demonstrating signal loss during acquisition, were excluded.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":3391825619,"compressed_storage_size":3392012771,"resource_type":"Database","access_policy":"Open","topics":["ecg","angiography"],"source_url":"https://physionet.org/content/staffiii/1.0.0/"},{"public_project_uuid":"8a083f98-5057-482f-8ef1-fd008fae0bed","slug":"picsdb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2QQ2M","is_latest_version":true,"short_description":"ECG and respiration recordings of ten preterm infants collected from a Neonatal Intensive Care Unit.","publish_date":"2017-02-09","title":"Preterm Infant Cardio-Respiratory Signals Database","abstract":"The Preterm Infant Cardio-respiratory Signals (PICS) database contains simultaneous ECG and respiration recordings of ten preterm infants collected from the Neonatal Intensive Care Unit (NICU) of the University of Massachusetts Memorial Healthcare. Statistical features based on linear estimates of heart rate are used to predict episodes of bradycardia.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1730761576,"compressed_storage_size":1730777744,"resource_type":"Database","access_policy":"Open","topics":["infant","ecg","multiparameter","neonate","respiration"],"source_url":"https://physionet.org/content/picsdb/1.0.0/"},{"public_project_uuid":"1a2b0bb3-2f8e-4d97-9b96-c496aa829ae4","slug":"taichidb","version":"1.0.0","core_doi":"10.13026/9v35-xa84","version_doi":"10.13026/C2W01J","is_latest_version":false,"short_description":"","publish_date":"2017-05-11","title":"Tai Chi, Physiological Complexity, and Healthy Aging - Gait","abstract":"The "Tai Chi, Physiological Complexity, and Healthy Aging—Gait" database contains two-channel gait recordings of 87 older adults from the Greater Boston area.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":527662636,"compressed_storage_size":527753482,"resource_type":"Database","access_policy":"Open","topics":["stability","gait"],"source_url":"https://physionet.org/content/taichidb/1.0.0/"},{"public_project_uuid":"5d2e5478-a396-45fd-b6e8-dddd6627a2f8","slug":"ltrsvp","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2KX0P","is_latest_version":true,"short_description":"","publish_date":"2017-05-19","title":"EEG Signals from an RSVP Task","abstract":"This project contains EEG data from 11 healthy participants upon rapid presentation of images through the Rapid Serial Visual Presentation (RSVP) protocol at speeds of 5, 6, and 10 Hz.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":785298399,"compressed_storage_size":785310619,"resource_type":"Database","access_policy":"Open","topics":["imaging","eeg","neuroelectric"],"source_url":"https://physionet.org/content/ltrsvp/1.0.0/"},{"public_project_uuid":"ffd4175b-edae-40b7-ac04-4c7f6129dd97","slug":"videopulse","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C28D5M","is_latest_version":true,"short_description":"Pulse signal recordings obtained from 15 healthy volunteers.","publish_date":"2017-07-12","title":"Video Pulse Signals in Stationary and Motion Conditions","abstract":"The data set contains 40 pulse signal recordings obtained from 15 healthy volunteers (12 men and 3 women between 23 and 35 years). For each subject, there is a simultaneous face video recording, and a finger pulse recording reference system, in a sitting position.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":5758358,"compressed_storage_size":5797220,"resource_type":"Database","access_policy":"Open","topics":["video","pulse"],"source_url":"https://physionet.org/content/videopulse/1.0.0/"},{"public_project_uuid":"e87c0127-b799-48d7-9b4d-dc62ffc83590","slug":"sufhsdb","version":"1.0.0","core_doi":"10.13026/9egj-jg49","version_doi":"10.13026/C2G66S","is_latest_version":false,"short_description":"","publish_date":"2017-07-19","title":"Shiraz University Fetal Heart Sounds Database","abstract":"The Shiraz University (SU) fetal heart sounds database (SUFHSDB) contains fetal and maternal PCG recordings from 109 pregnant women.
\r\n\r\n","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":827597838,"compressed_storage_size":827680924,"resource_type":"Database","access_policy":"Open","topics":["fetal","synthetic","phonocardiography","pregnancy"],"source_url":"https://physionet.org/content/sufhsdb/1.0.0/"},{"public_project_uuid":"1050b4ee-2ffa-41cc-9fd1-0faaff5f9c1d","slug":"noneeg","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C26Q2Z","is_latest_version":true,"short_description":"Non-EEG physiological signals collected using non-invasive wrist worn biosensors and consists of electrodermal activity, temperature, acceleration, heart rate, and arterial oxygen level.","publish_date":"2017-07-19","title":"Non-EEG Dataset for Assessment of Neurological Status","abstract":"
This database contains non-EEG physiological signals collected at Quality of Life Laboratory at University of Texas at Dallas, used to infer the neurological status (including physical stress, cognitive stress, emotional stress and relaxation) of 20 healthy subjects. The data was collected using non-invasive wrist worn biosensors and consists of electrodermal activity (EDA), temperature, acceleration, heart rate (HR), and arterial oxygen level (SpO2).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":3900280,"compressed_storage_size":3925168,"resource_type":"Database","access_policy":"Open","topics":["temperature","acceleration","multiparameter","heart rate","electrodermal activity"],"source_url":"https://physionet.org/content/noneeg/1.0.0/"},{"public_project_uuid":"940c8462-f648-445b-90e5-731f5c1baafb","slug":"unicaprop","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2BH3H","is_latest_version":true,"short_description":"Differential biopotential measurements recorded from the tongues of 39 healthy voluntary human subjects.","publish_date":"2017-07-19","title":"UniCA ElectroTastegram Database (PROP)","abstract":"The UniCA ElectroTastegram Database (PROP) contains 39 differential biopotential measurements recorded from the tongues of as many healthy voluntary human subjects (16 males, 23 females, equally divided into the three PROP taster status classes), during a stimulation with 30uL, 3.2 mmol/L solution of 6-n-propylthiouracil (PROP).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":8317174,"compressed_storage_size":8331254,"resource_type":"Database","access_policy":"Open","topics":["taste","neuroelectric","tongue"],"source_url":"https://physionet.org/content/unicaprop/1.0.0/"},{"public_project_uuid":"def79edc-aa54-4edc-8b75-4f0527637a59","slug":"qde","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C23082","is_latest_version":true,"short_description":"","publish_date":"2017-08-10","title":"Quantitative Dehydration Estimation","abstract":"Quantitative estimation of dehydration (total body water loss) using bioimpedance measurements, temperature measurements, salivary samples, and sweat samples.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":14511,"compressed_storage_size":14933,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/qde/1.0.0/"},{"public_project_uuid":"1544d859-33a1-432f-a45e-33483fe33079","slug":"earh","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2Z664","is_latest_version":true,"short_description":"Auditory Brainstem Response and Otoacoustic Emission recordings in eight hearing impaired listeners.","publish_date":"2017-09-26","title":"Evoked Auditory Responses in Heading Impaired","abstract":"The Evoked Auditory Responses in Hearing Impaired database contains evoked Auditory Brainstem Response (ABR) and Otoacoustic Emission (OAE) recordings in eight hearing impaired listeners, in response to tone-burst stimuli across a wide range of levels.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":14001565408,"compressed_storage_size":14001835974,"resource_type":"Database","access_policy":"Open","topics":["sound","auditory","deafness","neuroelectric"],"source_url":"https://physionet.org/content/earh/1.0.0/"},{"public_project_uuid":"7f725379-b265-4504-a3c1-1a0010a6143e","slug":"osv","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2TH47","is_latest_version":true,"short_description":"","publish_date":"2017-09-27","title":"Pattern Analysis of Oxygen Saturation Variability","abstract":"This database contains one hour oxygen saturation measurements of 36 patients, used for the analysis of oxygen saturation variability.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":992496,"compressed_storage_size":1016054,"resource_type":"Database","access_policy":"Open","topics":["oxygen saturation","oximetry","cardiovascular"],"source_url":"https://physionet.org/content/osv/1.0.0/"},{"public_project_uuid":"ab96d696-d90a-4890-b8e4-92ede1e416a5","slug":"pmd","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2W09K","is_latest_version":true,"short_description":"Pressure sensor data captured from 13 participants in various sleeping postures.","publish_date":"2017-09-28","title":"A Pressure Map Dataset for In-bed Posture Classification","abstract":"In this project, we collected in-bed posture pressure data from multiple adult participants using two different types of pressure sensing mats. To the best of our knowledge, our dataset, PmatData is the first publicly-available dataset of pressure sensor data which includes various sleeping postures. PmatData contains pressure data from two separate experiments.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":107124728,"compressed_storage_size":107315828,"resource_type":"Database","access_policy":"Open","topics":["sleep","pressure","posture","gait"],"source_url":"https://physionet.org/content/pmd/1.0.0/"},{"public_project_uuid":"74dad065-1563-4493-b4ca-2b162d2424d1","slug":"tappy","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2K08D","is_latest_version":true,"short_description":"","publish_date":"2017-10-20","title":"Tappy Keystroke Data","abstract":"This is the keystroke dataset for the study titled 'High-accuracy detection of early Parkinson's Disease using multiple characteristics of finger movement while typing'. This research report is currently under review for publication by PLOS ONE.
\r\n\r\nThe dataset contains keystroke logs collected from over 200 subjects, with and without Parkinson's Disease (PD), as they typed normally on their own computer (without any supervision) over a period of weeks or months (having initially installed a custom keystroke recording app, Tappy). This dataset has been collected and analyzed in order to indicate that the routine interaction with computer keyboards can be used to detect changes in the characteristics of finger movement in the early stages of PD.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":89249592,"compressed_storage_size":89250102,"resource_type":"Database","access_policy":"Open","topics":["movement","parkinsons","neuroelectric"],"source_url":"https://physionet.org/content/tappy/1.0.0/"},{"public_project_uuid":"128b04cd-6dc1-4223-b380-baecdcf5a02c","slug":"wrist","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2PQ1X","is_latest_version":true,"short_description":"Photoplethysmogram recorded from 8 volunteers during walking, running and bike riding.","publish_date":"2017-10-20","title":"Wrist PPG During Exercise","abstract":"This database contains wrist PPGs recorded during walking, running and bike riding. Simultaneous motion estimates are collected using both accelerometers and gyroscopes to give multiple options for the removal of motion interference from the PPG traces. A reference chest ECG is included to allow a gold-standard comparison of heart rate during exercise.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":52170427,"compressed_storage_size":52180873,"resource_type":"Database","access_policy":"Open","topics":["movement","ecg","photoplethysmogram","accelerometer","multiparameter"],"source_url":"https://physionet.org/content/wrist/1.0.0/"},{"public_project_uuid":"df2b7091-735a-4d43-98fa-e0d24efb393d","slug":"pwave","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2108F","is_latest_version":true,"short_description":"P-wave annotations for twelve signals from the MIT-BIH Arrhythmia Database.","publish_date":"2018-01-09","title":"MIT-BIH Arrhythmia Database P-Wave Annotations","abstract":"This database contains reference P-wave annotations for twelve signals from the MIT-BIH Arrhythmia Database.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":23450466,"compressed_storage_size":23458200,"resource_type":"Database","access_policy":"Open","topics":["ecg","arrhythmia"],"source_url":"https://physionet.org/content/pwave/1.0.0/"},{"public_project_uuid":"d21f869a-48d5-4439-bd60-c2cf1fab6979","slug":"cuiless16","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2F66G","is_latest_version":true,"short_description":"A corpus of Concept Unique Identifier concepts taken from the SemEval2015 Task 14.","publish_date":"2018-01-24","title":"CUILESS2016","abstract":"The Concept Unique Identifier (CUI)-less database contains a corpus of "CUI-less" concepts taken from the SemEval2015 Task 14 that have been assigned CUIs. The annotation process allows assignment of CUIS from any Unified Medical Language System (UMLS) semantic group and compositional normalization using more than one CUI per disease entity. Concepts are mapped to SNOMED CT as represented in the September 2016 version found in Unified Medical Language System (UMLS) 2016AB.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":151893,"compressed_storage_size":152539,"resource_type":"Database","access_policy":"Open","topics":["concept","umls","snomed"],"source_url":"https://physionet.org/content/cuiless16/1.0.0/"},{"public_project_uuid":"4a0d0848-d987-4c1c-b5b3-0b0190eda4f6","slug":"sleepbrl","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C29H4K","is_latest_version":true,"short_description":"","publish_date":"2018-02-07","title":"Sleep Bioradiolocation Database","abstract":"The database contains 32 records of non-contact sleep monitoring by a bioradar. The records are accompanied by results of sleep scoring, based on polysomnography according to the rules of the American Academy of Sleep Medicine.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1630834846,"compressed_storage_size":1630846528,"resource_type":"Database","access_policy":"Open","topics":["sleep","bioradar","polysomnography"],"source_url":"https://physionet.org/content/sleepbrl/1.0.0/"},{"public_project_uuid":"478881c2-0335-4d98-bce8-ea5de091acd6","slug":"voiced","version":"1.0.0","core_doi":"10.13026/twfd-kb89","version_doi":"10.13026/C25Q2N","is_latest_version":true,"short_description":"","publish_date":"2018-06-07","title":"VOICED Database","abstract":"This database includes 208 voice samples, from 150 pathological, and 58 healthy voices.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":115343616,"compressed_storage_size":115492496,"resource_type":"Database","access_policy":"Open","topics":["sound","voice","auditory"],"source_url":"https://physionet.org/content/voiced/1.0.0/"},{"public_project_uuid":"dd6d27d3-2b55-4ab6-9228-37196cfe649f","slug":"bidmc","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2208R","is_latest_version":true,"short_description":"ECG signals extracted from the MIMIC-II Matched Waveform Database, with manual breath annotations added by annotators using impedance respiratory signal.","publish_date":"2018-06-20","title":"BIDMC PPG and Respiration Dataset","abstract":"This dataset contains signals and numerics extracted from the much larger MIMIC II matched waveform Database, along with manual breath annotations made from two annotators, using the impedance respiratory signal.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":217807814,"compressed_storage_size":217902224,"resource_type":"Database","access_policy":"Open","topics":["photoplethysmogram","multiparameter","ecg"],"source_url":"https://physionet.org/content/bidmc/1.0.0/"},{"public_project_uuid":"c8f03341-f35b-414a-b712-2b663aaa7788","slug":"plantar","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C20Q1Q","is_latest_version":true,"short_description":"Plantar pressure distribution and muscle activity during gait from 20 healthy male adults.","publish_date":"2018-07-17","title":"Modulation of Plantar Pressure and Muscle During Gait","abstract":"Modulation of plantar pressure and muscle during gait was acquired to investigate the influences of electrical stimulation (ES) of the tibialis anterior (TA) on plantar pressure and the gastrocnemius medialis (GM) activity during gait in healthy adults. The database consists of plantar pressure distribution and two muscles activity during gait from 20 healthy male adults. We evaluated the activities of gastrocnemius medialis and lateralis during gait in the preliminary study, and observed that gastrocnemius medialis is more sensitive to gait. Therefore, we recorded GM and TA in this study. Only subjects who do not have any history of either neurological or orthopedic impairments were included..
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":18712036,"compressed_storage_size":18764944,"resource_type":"Database","access_policy":"Open","topics":["pressure","gait"],"source_url":"https://physionet.org/content/plantar/1.0.0/"},{"public_project_uuid":"eae44fc3-9541-413c-b096-a4a9c08d7bc4","slug":"tpehgt","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2166R","is_latest_version":true,"short_description":"Electrohysterogram signals accompanied by a simultaneously recorded external tocogram.","publish_date":"2018-08-29","title":"Term-Preterm EHG DataSet with Tocogram","abstract":"The Term-Preterm ElectroHysteroGram DataSet with Tocogram (TPEHGT DS) contains 26 four-signal 30-min uterine EHG records, i.e., three EHG signals accompanied by a simultaneously recorded external tocogram measuring mechanical uterine activity (TOCO signal) of pregnant women, and another five 30-min uterine records (EHG signals and TOCO signal) of non-pregnant women.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":41759705,"compressed_storage_size":41778923,"resource_type":"Database","access_policy":"Open","topics":["neuroelectric","pregnancy","electrohysterogram"],"source_url":"https://physionet.org/content/tpehgt/1.0.0/"},{"public_project_uuid":"3da272cf-a80b-4391-8696-7f2b8d5ea476","slug":"culm","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C24T0M","is_latest_version":true,"short_description":"Hand trajectory data collected from ten subjects as they performed various upper-limb motor tasks.","publish_date":"2018-08-30","title":"Complex Upper-Limb Movements","abstract":"The Complex Upper-Limb Movements database contains hand trajectory data collected from ten subjects as they performed various upper-limb motor tasks. The data was used in the above-referenced manuscript to identify the motor primitives contributing to the observed motor patterns.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":57334742,"compressed_storage_size":57423840,"resource_type":"Database","access_policy":"Open","topics":["movement"],"source_url":"https://physionet.org/content/culm/1.0.0/"},{"public_project_uuid":"8e049609-4c32-4b21-b596-a1673a8a7207","slug":"cves","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2DW96","is_latest_version":true,"short_description":"Multimodal data from a large study investigating the effects of ischemic stroke on cerebral vasoregulation.","publish_date":"2018-10-04","title":"Cerebral Vasoregulation in Elderly with Stroke","abstract":"This database contains multimodal data from a large study investigating the effects of ischemic stroke on cerebral vasoregulation. The cross sectional study compared 60 subjects who suffered strokes, to 60 control subjects, collecting the following data for each patient across multiple days: transcranial doppler of cerebral arteries, 24-h blood pressure numerics, high resolution waveforms (ECG, blood pressure, CO2 and respiration) during various movement tasks, 24-h ECG, EMG, and accelerometer recordings, and gait pressure recordings during a walking test.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":186701850685,"compressed_storage_size":186702341356,"resource_type":"Database","access_policy":"Open","topics":["stroke","multiparameter","vasoregulation"],"source_url":"https://physionet.org/content/cves/1.0.0/"},{"public_project_uuid":"0524894f-eaed-4a46-b17a-9e12bd0429a4","slug":"rvmh1","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C24M3P","is_latest_version":true,"short_description":"","publish_date":"2018-11-14","title":"Response to Valsalva Maneuver in Humans","abstract":"Functional metrics of autonomic control of heart rate, including baroreflex sensitivity, have been shown to be strongly associated with cardiovascular risk. A decrease in baroreflex sensitivity with aging is hypothesized to represent a contributing causal factor in the etiology of primary hypertension. To assess baroreflex function in human subjects, two complementary methods to simulate the response in heart rate elicited by the Valsalva maneuver were developed and applied to data obtained from a cohort of healthy normal volunteers.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":6324934,"compressed_storage_size":6346594,"resource_type":"Database","access_policy":"Open","topics":["valsalva","multiparameter","heart rate"],"source_url":"https://physionet.org/content/rvmh1/1.0.0/"},{"public_project_uuid":"148efb25-bdc9-422e-a1c0-5321eb51f8eb","slug":"ecgcipa","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2967M","is_latest_version":true,"short_description":"ECG effects of ranolazine, verapamil, lopinavir+ritonavir, chloroquine, dofetilide, diltiazem, and dofetilide+diltiazem in a small clinical study.","publish_date":"2018-12-13","title":"CiPA ECG Validation Study","abstract":"ECG effects of ranolazine, verapamil, lopinavir+ritonavir, chloroquine, dofetilide, diltiazem, and dofetilide+diltiazem in a small sample size clinical study. The ECGCIPA database contains multi-channel ECG recordings of 60 subjects participating in the CiPA ECG validation study.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":3250469043,"compressed_storage_size":3257689247,"resource_type":"Database","access_policy":"Open","topics":["medication","ecg"],"source_url":"https://physionet.org/content/ecgcipa/1.0.0/"},{"public_project_uuid":"fd4c3c48-4fcb-426a-9873-527f70fd75ca","slug":"eegmat","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2JQ1P","is_latest_version":true,"short_description":"","publish_date":"2018-12-17","title":"EEG During Mental Arithmetic Tasks","abstract":"The database contains EEG recordings of subjects before and during the performance of mental arithmetic tasks.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":183620001,"compressed_storage_size":183634285,"resource_type":"Database","access_policy":"Open","topics":["eeg","neuroelectric"],"source_url":"https://physionet.org/content/eegmat/1.0.0/"},{"public_project_uuid":"a5da51be-cd60-4055-b3f3-1f08c1b3e62e","slug":"wctecgdb","version":"1.0.0","core_doi":"10.13026/2mz7-3v18","version_doi":"10.13026/wrn5-3v80","is_latest_version":false,"short_description":"","publish_date":"2019-03-08","title":"Wilson Central Terminal ECG Database","abstract":"We present a unique dataset which contains the Wilson Central Terminal (WCT), the voltage of right arm (RA), left arm (LA) and left leg (LL), six unipolar chest leads associated with three limb leads and six precordial leads.
\r\n\r\nWith our ECG device, the potentials of RA, LA, LL and six true unipolar leads are measured using the right leg as the reference point. Having this configuration, we have been able to overcome the difficulty of recording the WCT in a clinical setting. Our data confirmed once again that the WCT does not have a small value, and could be as high as 241% of lead II and has standard ECG characteristics such as the p-wave and the t-wave. The dataset comprises of 540 ten second segments recorded from 92 patients.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":321040832,"compressed_storage_size":321259564,"resource_type":"Database","access_policy":"Open","topics":["electrocardiography","wilson central terminal","limb potential","ecg"],"source_url":"https://physionet.org/content/wctecgdb/1.0.0/"},{"public_project_uuid":"1b68216e-8d61-46df-a9b2-ed6fe3e0c38d","slug":"phdsm","version":"1.0","core_doi":"10.13026/5889-jj86","version_doi":"10.13026/C23H3B","is_latest_version":false,"short_description":"","publish_date":"2019-03-15","title":"Permittivity of Healthy and Diseased Skeletal Muscle","abstract":"A better understanding of the permittivity of skeletal muscle is essential for the development of new diagnostic tools and approaches for neuromuscular evaluation. However, there remain important knowledge gaps in our understanding of this property in healthy and diseased skeletal muscle, which hinder its translation into clinical application. The aim of this project is to provide a platform for the researchers to contribute and share the dissemination of this property in healthy and diseased muscle. Ultimately, the normative data reported will offer the scientific community the opportunity to improve the accuracy of existing techniques as well as developing new diagnostic tools and therapies.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":74182,"compressed_storage_size":76612,"resource_type":"Database","access_policy":"Open","topics":["skeletal muscle","permittivity","anisotropy","neuromuscular disorders"],"source_url":"https://physionet.org/content/phdsm/1.0/"},{"public_project_uuid":"723de0dd-fa21-420b-926a-088b2da33034","slug":"mimiciii","version":"1.4","core_doi":"10.13026/cd7z-wg25","version_doi":"10.13026/C2XW26","is_latest_version":true,"short_description":"","publish_date":"2016-09-04","title":"MIMIC-III Clinical Database","abstract":"MIMIC-III is a large, freely-available database comprising deidentified health-related data associated with over forty thousand patients who stayed in critical care units of the Beth Israel Deaconess Medical Center between 2001 and 2012. The database includes information such as demographics, vital sign measurements made at the bedside (~1 data point per hour), laboratory test results, procedures, medications, caregiver notes, imaging reports, and mortality (including post-hospital discharge).
MIMIC supports a diverse range of analytic studies spanning epidemiology, clinical decision-rule improvement, and electronic tool development. It is notable for three factors: it is freely available to researchers worldwide; it encompasses a diverse and very large population of ICU patients; and it contains highly granular data, including vital signs, laboratory results, and medications.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":6620705643,"compressed_storage_size":6620711553,"resource_type":"Database","access_policy":"Credentialed","topics":["critical care","machine learning","natural language processing","intensive care","clinical"],"source_url":"https://physionet.org/content/mimiciii/1.4/"},{"public_project_uuid":"2c29165a-f4a4-4fd2-b9a8-e4dfb02ef7f2","slug":"inipdmsa","version":"1.0","core_doi":"10.13026/2572-db54","version_doi":"10.13026/bn1x-my19","is_latest_version":true,"short_description":"Dataset collected as part of a study that aimed to determine the effects of intranasal insulin on cognition and motor performance in Parkinson's disease.","publish_date":"2019-04-10","title":"Safety and Preliminary Efficacy of Intranasal Insulin for Cognitive Impairment in Parkinson Disease and Multiple System Atrophy","abstract":"This dataset was collected as part of a study that aimed to determine the effects of intranasal insulin (INI) on cognition and motor performance in PD. The study was a proof of concept, randomized, double-blinded, placebo-controlled trial evaluating the effects of 40 international units (IU) of insulin or saline once daily for four weeks on cognitive and functional performance. Of 16 subjects enrolled, eight in the INI group and six in the placebo group completed verbal fluency (FAS), Unified Parkinson Disease Scale (UPDRS), and modified Hoehn and Yahr scale (HY, PD severity) at baseline and post-treatment and were included in the analyses.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":7895456,"compressed_storage_size":7907010,"resource_type":"Database","access_policy":"Open","topics":["parkinsons","atrophy","insulin","randomized controlled trial","cognitive assessment","motor assessment"],"source_url":"https://physionet.org/content/inipdmsa/1.0/"},{"public_project_uuid":"79f8e045-78a1-48d4-84fc-567b18f95f6c","slug":"wfdb-swig-matlab","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C28378","is_latest_version":true,"short_description":"","publish_date":"2009-08-17","title":"WFDB SWIG Toolbox for MATLAB","abstract":"The WFDB SWIG Toolbox for MATLAB is a collection of applications for reading, writing, manipulating, and plotting PhysioBank data, implemented as functions in MATLAB.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":43715077,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["wfdb"],"source_url":"https://physionet.org/content/wfdb-swig-matlab/1.0.0/"},{"public_project_uuid":"21b80f71-ae78-4083-9160-dfec4710ce61","slug":"semia","version":"3.0.1","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2002-10-30","title":"SEMIA","abstract":"SEMIA, version 3.0.1, is a tool for viewing time series of diagnostic and morphology parameters of long-term ambulatory recordings, and ST segment annotations with their corresponding ECG waveforms of the Long-Term ST Database (LTST DB). SEMIA was created during the development of the LTST DB, a project supported by Medtronic, Inc. (Minneapolis, MN, USA) and Zymed, Inc. (Camarillo, CA, USA).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":710481,"resource_type":"Software","access_policy":"Open","topics":["multiparameter","visualization"],"source_url":"https://physionet.org/content/semia/3.0.1/"},{"public_project_uuid":"228d8bfa-d74f-4d1f-87bb-da1a68ac85c9","slug":"wfdb-swig","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2006-03-12","title":"WFDB SWIG","abstract":"The wfdb-swig package provides interfaces between the WFDB library and software written in a variety of other languages, so that such software can have access to the full range of capabilities supported by current and future versions of the WFDB library. The package provides interfaces for software written using Perl, Python, C# (and other .NET languages), and Java. The wfdb-swig package can be readily extended for use with other target languages supported by SWIG, such as PHP, Ruby, TCL, and several versions of Lisp.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":565078,"compressed_storage_size":582182,"resource_type":"Software","access_policy":"Open","topics":["wfdb"],"source_url":"https://physionet.org/content/wfdb-swig/1.0.0/"},{"public_project_uuid":"035b1fed-98c9-45da-b1d7-07da892a336d","slug":"dchromatix","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2015-11-09","title":"Data Chromatix","abstract":"Analysis of biomedical time series plays a key role in clinical management and basic investigation. However, most conventional monitors streaming data in real-time show only the most recent values, not referenced to past dynamics. The proposed visualization method (termed “data chromatix”) was developed to address this challenge by bringing memory of the system’s past behavior into the current display window.
\r\n\r\nThe function DataChromatix.m (version 1.0) assigns a color to each data point of a time series. The color is determined by the values of a normalized histogram (estimated probability density function) computed from a pre-selected segment of the data. The algorithm receives the time series as input and generates a video of its colorized version as it would look on a typical monitor display, as well as a static graph of the entire colorized signal.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["visualization"],"source_url":"https://physionet.org/content/dchromatix/1.0.0/"},{"public_project_uuid":"58df2c8b-e4db-4d78-83bf-76b805752dad","slug":"mpp","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2016-04-12","title":"Multiscale Poincare Plots","abstract":"Poincaré delay maps are widely used in the analysis of cardiac interbeat interval (RR) dynamics. To facilitate visualization of the structure of these complex time series, we introduced a methodology called multiscale Poincaré (MSP) plots.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["poincare","visualization"],"source_url":"https://physionet.org/content/mpp/1.0.0/"},{"public_project_uuid":"3f368c18-7341-446e-9f05-385eca428fc5","slug":"plt","version":"2.5","core_doi":null,"version_doi":"10.13026/C2T08K","is_latest_version":true,"short_description":"","publish_date":"2002-11-07","title":"plt - Software for 2D Plots","abstract":"plt is a non-interactive plotting utility originally written for Unix by Paul Albrecht. plt can produce publication-quality 2D plots in PostScript from easily-produced text or binary data files, and can also create screen plots under the X Window System.
record is an MS-DOS application that determines what signals and measurements are available from a Hewlett-Packard (Agilent/Philips) Merlin (a.k.a. CMS or Component Monitoring System) patient monitor, and then logs all of the available data to disk continuously until it is stopped by the user.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["monitor","mimic"],"source_url":"https://physionet.org/content/record/1.0.0/"},{"public_project_uuid":"89ae214e-0430-42b3-805d-c47a06c11061","slug":"deid","version":"1.1","core_doi":null,"version_doi":"10.13026/C20M3F","is_latest_version":true,"short_description":"","publish_date":"2007-12-18","title":"De-Identification Software Package","abstract":"The deid software package includes code and dictionaries for automated location and removal of protected health information (PHI) in free text from medical records.
edf-anonymize reads an EDF or EDF+ file (input), writing an anonymized copy of it as output.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["anonymization","deidentification","phi"],"source_url":"https://physionet.org/content/edf-anonymize/1.0.0/"},{"public_project_uuid":"4e248037-7449-477f-9545-5c511cf549d2","slug":"apdet","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2002-02-04","title":"Apnea Detection from the ECG","abstract":"Hilbert Transform based Sleep Apnea Detection using a Single Lead Electrocardiogram.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["sleep","ecg","apnea"],"source_url":"https://physionet.org/content/apdet/1.0.0/"},{"public_project_uuid":"2089f4f2-906b-4bc7-a6f7-14056de2e904","slug":"charis-gui","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2016-12-16","title":"Cerebral Haemodynamic Autoregulatory Information System GUI","abstract":"Acute Brain injury (ABI) is a devastating event requiring intensive acute treatment and post-injury rehabilitation, both delivered for indeterminate periods of time. For severe ABIs, acute treatment is aimed at stabilizing the patient to prevent secondary brain injury from ischemia and swelling. This requires a balance between adequate levels of cerebral blood flow and safely low intracranial pressure (ICP) - a task normally done by autoregulatory (AR) processes of the brain.
\n\nIn the absence of normal AR, hemodynamic stability is difficult to maintain because there are no reliable predictors to guide treatment. Thus, guidelines for triage and discharge are somewhat arbitrary for ABI and the need for a Clinical Decision Support (CDS) system for neurotrauma is widely recognized. Prerequisite to a CDS is a large database of patient records and efficient means to extract meaningful information from them.
\n\nCHARIS will systematize the analysis of relevant physiological signals, and will embody data-driven algorithms to search for potential predictors of acute clinical events.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["brain injury","intracranial pressure"],"source_url":"https://physionet.org/content/charis-gui/1.0.0/"},{"public_project_uuid":"47d4a117-dbbe-470f-8231-04c744cf9892","slug":"ecgpuwave","version":"1.3.4","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2018-10-29","title":"ECGPUWAVE","abstract":"ecgpuwave analyses an ECG signal from the specified record, detecting the QRS complexes and locating the beginning, peak, and end of the P, QRS, and ST-T waveforms. The output of ecgpuwave is written as a standard WFDB-format annotation file associated with the specified annotator.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["ecg"],"source_url":"https://physionet.org/content/ecgpuwave/1.3.4/"},{"public_project_uuid":"d44a8422-3f99-4814-adec-0cdae09296bd","slug":"cardiac-output","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2007-08-02","title":"Cardiac Output Estimation from Arterial Blood Pressure Waveforms","abstract":"This project contains the code used for a study that rigorously compares 11 estimates against thermodilution cardiac output (TCO) using 120 subjects from the MIMIC II database (all subjects for whom multiple TCO measurements were recorded at the time of the study).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["cardiac output","estimation"],"source_url":"https://physionet.org/content/cardiac-output/1.0.0/"},{"public_project_uuid":"7a41b421-24a0-4c27-b1a8-bd93d95df3f1","slug":"edr","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2000-03-19","title":"ECG-Derived Respiration","abstract":"The ECG-Derived Respiration (EDR) technique is an approach to obtain a respiration signal from an ECG.. It is based on the observation that the positions of ECG electrodes on the chest surface move relative to the heart, and transthoracic impedance varies, as the lungs fill and empty. Thus the lead axes vary at different points in the respiratory cycle, and any sufficiently precise measurement of the mean cardiac electrical axis shows variations that are correlated with respiration. This method is particularly effective if two or more ECG signals are available, but a single lead is sufficient. The EDR can be obtained even in the context of congestive heart failure, in which RSA may be absent.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["respiration"],"source_url":"https://physionet.org/content/edr/1.0.0/"},{"public_project_uuid":"d4e36036-da87-4d9e-9465-3fb3e3e8e282","slug":"puka","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2004-12-15","title":"Puka - Software for Detection of Breaths in Strain Gauge Recordings","abstract":"Puka calculates descriptive statistics such as heart rate variability (HRV), peak-valley respiratory sinus arrhythmia (RSA), and respiratory variables from EKG and strain gauge respiration data.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/puka/1.0.0/"},{"public_project_uuid":"6fc7e65e-2d28-4dbb-9ccc-0b51146c0678","slug":"twanalyser","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2008-11-18","title":"TWAnalyser - A T-wave Alternans Detector","abstract":"This software suite consists of Matlab implementations of the widely used Spectral Method (SM) and Modified Moving Average (MMA) for T-wave alternans (TWA) detection. It was initially developed for the PhysioNet/CinC Challenge 2008, in which these and a wide variety of other methods were evaluated and compared in blinded tests, using a data set that included ECGs with and without TWA as well as synthetic ECGs with calibrated amounts of artificial TWA. The implementation of SM presented here scored second in the open-source category and third overall with a score of 0.881. TWAanalyser's implementation of MMA received a score of 0.400. (A third method, which combined both algorithms, scored 0.834, not quite as well as SM alone. In this case, ECGs in which SM did not detect TWA were assigned a TWA estimate of 0, and all others were assigned the MMA algorithm's TWA estimate.) Most of the work to improve the score resulted from adjustments to filtering and alignment.
\n\nTWAnalyser has been tested on both Windows XP (SP2) and Linux (Fedora Core 4-9) using Matlab R13 (6.5) and R2006b (7.3).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["ecg"],"source_url":"https://physionet.org/content/twanalyser/1.0.0/"},{"public_project_uuid":"5f41a6bb-275c-41d1-84f5-567ad7f14834","slug":"ecgkit","version":"1.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2015-11-09","title":"ECG-Kit","abstract":"This toolbox is a collection of Matlab tools that Mariano Llamedo Soria used, adapted or developed during his PhD and post-doc work with the Besicos group at University of Zaragoza, Spain and at the National Technological University of Buenos Aires, Argentina.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["wfdb","ecg"],"source_url":"https://physionet.org/content/ecgkit/1.0/"},{"public_project_uuid":"bdc2b8cc-1977-43f9-9547-d2f13d0e8f1f","slug":"geh","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2018-05-02","title":"Measurement of Global Electrical Heterogeneity","abstract":"The Global Electrical Heterogeneity (GEH) concept is based on the theory of Wilson’s electrical gradient vector, which characterizes the degree of heterogeneity of the total recovery time across the ventricles.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/geh/1.0.0/"},{"public_project_uuid":"c16fdf28-013c-4389-87f0-ba8c99bc6a9b","slug":"dddmaps","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2015-01-31","title":"Dynamical Density Delay Maps","abstract":"This project contains Matlab functions for visualizing the behavior of complex systems by means of the dynamical density delay mapping ("D3-Map") technique . This simple-to-implement visualization method provides an animated representation of a system's dynamics.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":49362030,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["visualization","complexity"],"source_url":"https://physionet.org/content/dddmaps/1.0.0/"},{"public_project_uuid":"4a4e5493-5c55-4ef5-8959-7df7f149a278","slug":"dfa","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2001-11-11","title":"Detrended Fluctuation Analysis","abstract":"The method of detrended fluctuation analysis has proven useful in revealing the extent of long-range correlations in time series.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":496104,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["complexity"],"source_url":"https://physionet.org/content/dfa/1.0.0/"},{"public_project_uuid":"12f9de74-a6f8-4ae8-b1ea-89c2089c51a1","slug":"gmse","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2019-02-02","title":"Generalized Multiscale Entropy Analysis","abstract":"The method of generalized multiscale entropy (GMSE) analysis is useful for investigating complexity in physiologic signals and other series that have correlations at multiple (time) scales. It represents a generalization of the original method of multiscale entropy analysis.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":2723060,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["entropy","complexity"],"source_url":"https://physionet.org/content/gmse/1.0.0/"},{"public_project_uuid":"2a1c4663-1c0e-4c3f-be0e-d301bc53c6c4","slug":"ibs","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2004-10-26","title":"Information-Based Similarity","abstract":"IBS is a novel method for comparing time series based on linguistic analysis of "word" frequency and rank order statistics. It has been applied successfully in a variety of applications, including analysis of interbeat (RR) intervals, classification of the SARS coronavirus, and study of literary works of disputed authorship.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":826616,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["interbeat","complexity","rr interval"],"source_url":"https://physionet.org/content/ibs/1.0.0/"},{"public_project_uuid":"818f2635-3214-4f20-932d-f68eaed77e40","slug":"lyapunov","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2001-01-16","title":"A practical method for calculating Lyapunov exponents from small data sets","abstract":"Detecting the presence of chaos in a dynamical system is an important problem that is solved by measuring the largest Lyapunov exponent. Lyapunov exponents quantify the exponential divergence of initially close state-space trajectories and estimate the amount of chaos in a system.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1468801,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["chaos","complexity"],"source_url":"https://physionet.org/content/lyapunov/1.0.0/"},{"public_project_uuid":"066d2d49-54c9-44bc-8728-2687490bc0ab","slug":"mma","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2015-01-31","title":"Multiscale Multifractal Analysis","abstract":"Multiscale multifractal analysis (MMA) (Gierałtowski et al., 2012) is a time series analysis method, designed to describe scaling properties of fluctuations within the signal analyzed. The main result of this procedure is the so called Hurst surface h(q,s) , which is a dependence of the local Hurst exponent h (fluctuation scaling exponent) on the multifractal parameter q (Kantelhardt et al., 2002) and the scale of observation s (data window width).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":2239654,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["entropy","complexity"],"source_url":"https://physionet.org/content/mma/1.0.0/"},{"public_project_uuid":"46a12f85-95ba-483b-a238-ed1c31d3fecc","slug":"mse","version":"1.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2004-08-04","title":"Multiscale Entropy Analysis","abstract":"The method of multiscale entropy (MSE) analysis is useful for investigating complexity in physiologic signals and other series that have correlations at multiple (time) scales.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":6837175,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["entropy","complexity"],"source_url":"https://physionet.org/content/mse/1.0/"},{"public_project_uuid":"be0a614c-eb76-478d-a491-f73dc8f37708","slug":"multifractal","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2002-05-29","title":"Software for Analysis of Multifractal Time Series","abstract":"Software for calculating multifractal partitions and moments of a time series.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":9221229,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["complexity"],"source_url":"https://physionet.org/content/multifractal/1.0.0/"},{"public_project_uuid":"b5396a72-b757-47e4-ac9e-fbd7c0f47905","slug":"sampen","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2004-11-08","title":"Sample Entropy Estimation","abstract":"Sample Entropy is a useful tool for investigating the dynamics of heart rate and other time series. Sample Entropy is the negative natural logarithm of an estimate of the conditional probability that subseries (epochs) of length m that match pointwise within a tolerance r also match at the next point.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":586472,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["entropy","complexity"],"source_url":"https://physionet.org/content/sampen/1.0.0/"},{"public_project_uuid":"00ff7b36-938a-4e66-a738-4e35d202dfd8","slug":"tewp","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2016-03-04","title":"Transfer Entropy With Partitioning","abstract":"This is a repository of MATLAB functions that can estimate transfer entropy (information flow) from one time series to another using a non-parametric partitioning algorithm.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":155718,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["entropy","complexity"],"source_url":"https://physionet.org/content/tewp/1.0.0/"},{"public_project_uuid":"6f78a9e2-0e85-41e9-94de-4a346b606b99","slug":"activity","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2002-04-18","title":"Estimating Activity from Instantaneous Heart Rate","abstract":"Using only a heart rate time series, it is possible to measure a number of features that reflect the level of physical activity. Here we provide software for deriving an "activity index" based on measurements of mean heart rate, total power of the instantaneous heart rate time series over a given interval, and stationarity.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":309408,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["activity","heart rate"],"source_url":"https://physionet.org/content/activity/1.0.0/"},{"public_project_uuid":"4977080c-ee15-4336-8178-843668560c3a","slug":"pnnx","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2003-03-03","title":"pNNx - Time Domain Heart Rate Variability Analysis","abstract":"The pNNx software package expands the analysis capabilities of the widely used pNN50 measure, providing researchers with the flexibility to test various thresholds for optimal discrimination of HRV in populations of interest.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":17370965,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["interbeat","hrv"],"source_url":"https://physionet.org/content/pnnx/1.0.0/"},{"public_project_uuid":"2e4371de-6b84-48d1-9e3d-86bf1c498ebc","slug":"heartprints","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2003-03-08","title":"Heartprints - A Dynamical Portrait of Cardiac Arrhythmia","abstract":"As part of an effort to visualize and to understand the dynamics of cardiac arrhythmias with frequent ventricular ectopy, we have developed a new representational technique termed "heartprints." These heartprints are a combination of seven histograms generated from an annotated record of heartbeat time intervals.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":100976,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["interbeat","rr interval","arrhythmia"],"source_url":"https://physionet.org/content/heartprints/1.0.0/"},{"public_project_uuid":"457e1526-3a5a-465f-a4a2-3a1bcd925154","slug":"cvsim","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2007-03-16","title":"CVSim","abstract":"CVSim is a lumped-parameter model of the human cardiovascular system that has been developed and used for research and for teaching quantitative physiology courses at MIT and Harvard Medical School since 1984. The versions presented here have a graphical user interface implemented in Java, with computational components in C.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":200251932,"compressed_storage_size":168639545,"resource_type":"Software","access_policy":"Open","topics":["simulation","cardiovascular"],"source_url":"https://physionet.org/content/cvsim/1.0.0/"},{"public_project_uuid":"47e4ff67-b834-4f03-85f9-e097ad155f3c","slug":"rcvsim","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2001-09-19","title":"A Cardiovascular Simulator for Research","abstract":"We introduce here the Research CardioVascular SIMulator (RCVSIM) software in order to complement research with the experimental data sets provided by PhysioBank. The human cardiovascular model upon which RCVSIM is based includes three major components. The first component is a lumped parameter model of the pulsatile heart and circulation. The second component is a short-term regulatory system model which includes an arterial baroreflex system, a cardiopulmonary baroreflex system, and a direct neural coupling mechanism between respiration and heart rate.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":113899261,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["simulation","cardiovascular"],"source_url":"https://physionet.org/content/rcvsim/1.0.0/"},{"public_project_uuid":"71a6754c-bbc1-4de8-8b55-99cf5bf19d57","slug":"afvp","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2007-08-09","title":"AFVP - A Realistic Ventricular Rhythm Model During AF","abstract":"AFVP generates a synthesized beat-to-beat interval sequence of ventricular excitations with a realistic structure observed during AF.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":97334,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["interbeat","simulation","rr interval"],"source_url":"https://physionet.org/content/afvp/1.0.0/"},{"public_project_uuid":"906bfd85-6e8c-4ebf-ad2e-5abda59a3270","slug":"ecgsyn","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2003-12-03","title":"ECGSYN - A realistic ECG waveform generator","abstract":"ECGSYN generates a synthesized ECG signal with user-settable mean heart rate, number of beats, sampling frequency, waveform morphology (P, Q, R, S, and T timing, amplitude,and duration), standard deviation of the RR interval, and LF/HF ratio (a measure of the relative contributions of the low and high frequency components of the RR time series to total heart rate variability).
","license":{"name":"GNU General Public License version 3"},"dua":null,"main_storage_size":3385206,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["generator","simulation","ecg"],"source_url":"https://physionet.org/content/ecgsyn/1.0.0/"},{"public_project_uuid":"38687720-cabd-4312-808d-fe73d90a5cf2","slug":"ecgwavegen","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2002-11-23","title":"ECG waveform generator for Matlab/Octave","abstract":"ECGwaveGen generates a synthesized ECG signal with user-settable heart rate, signal duration, sampling frequency, QRS amplitude and duration, and T-wave amplitude; it uses QRSpulse to create premature beats followed by compensatory pauses.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":13172,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["generator","waveform","ecg"],"source_url":"https://physionet.org/content/ecgwavegen/1.0.0/"},{"public_project_uuid":"b7a0f988-4e97-471f-8813-7ae53a4b1a91","slug":"fecgsyn","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2014-11-04","title":"FECGSYN Toolbox","abstract":"The FECGSYN toolbox is a reference open-source platform for NI-FECG research, product of a collaboration between the Department of Engineering Science, University of Oxford (DES-OX), the Institute of Biomedical Engineering, TU Dresden (IBMT-TUD) and the Biomedical Engineering Faculty at the Technion Israel Institute of Technology (BME-IIT).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":79557161,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["fetal","ecg"],"source_url":"https://physionet.org/content/fecgsyn/1.0.0/"},{"public_project_uuid":"58225fff-4363-4d7f-8216-63f3386f47b0","slug":"gradient-algorithm","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2015-08-21","title":"Gradient Algorithm","abstract":"In finding an optimal stimulus waveform for inducing switches in neuronal states, analytical techniques from optimal control theory are often found to be difficult to use or extremely time intensive. Here, we present the code for a gradient-based algorithm approach that has been used to find energetically optimal stimulus waveforms to trigger an action potential in the Hodgkin-Huxley model as well as initiating and repressing repetitive firing in the FitzHugh-Nagumo models. These two models serve just as examples and the code can be easily adapted to any other system.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":2516526,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["optimization"],"source_url":"https://physionet.org/content/gradient-algorithm/1.0.0/"},{"public_project_uuid":"775eb1ba-05eb-4677-bafc-d621bb271827","slug":"eval-st","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2495B","is_latest_version":true,"short_description":"","publish_date":"2004-09-17","title":"EVAL_ST Tool","abstract":"The EVAL_ST tool is an open source tool to evaluate and compare performance and robustness of ST episode detection algorithms. The tool supports all standard and other relevant performance measures, aggregate gross and average statistics, and bootstrap statistical procedure to predict real-world clinical performance. The tool (written in C) is compilable an a wide variety of platforms and contains an additional graphic user interface module (LessTif/Motif environment) for use on the LINUX/UNIX operating systems.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":10441373,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["arrhythmia"],"source_url":"https://physionet.org/content/eval-st/1.0.0/"},{"public_project_uuid":"b411eb9e-852c-4ca2-bf5f-2d1c54c4ab03","slug":"random-search","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2014-11-04","title":"Random Search Toolbox","abstract":"A major issue with many signal processing and machine learning algorithms is the lack of optimisation methods for determining the numerous hyper-parameters associated with the model as well as the knowledge of which hyper-parameters are relevant. These parameters are usually tunned by trial and error (manual search) or by grid search. In 2012 Bergstra et al. showed empirically that random search was performing similar or better than grid search while reducing the computational cost substantially.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1415730,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/random-search/1.0.0/"},{"public_project_uuid":"0ec7e84c-32f5-4343-ac93-d37037f9bb51","slug":"pcst","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2018-04-28","title":"PhysioNet Cardiovascular Signal Toolbox","abstract":"The PhysioNet Cardiovascular Signal Toolbox is an open-source modular program for calculating heart rate variability (HRV) implemented in Matlab with evidence-based algorithms and output formats. The Toolbox is compatible with 64-bit MATLAB on GNU/Linux, Mac OS X, and MS-Windows.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":52409280,"compressed_storage_size":0,"resource_type":"Software","access_policy":"Open","topics":["heart rate variability","cardiovascular"],"source_url":"https://physionet.org/content/pcst/1.0.0/"},{"public_project_uuid":"efeee4d0-9799-4f09-8c86-43665f9fe59c","slug":"eicu-crd","version":"2.0","core_doi":"10.13026/0pzc-dm64","version_doi":"10.13026/C2WM1R","is_latest_version":true,"short_description":"Multi-center database comprising deidentified health data associated with over 200,000 admissions to ICUs across the United States between 2014-2015.","publish_date":"2019-04-15","title":"eICU Collaborative Research Database","abstract":"The eICU Collaborative Research Database is a multi-center database comprising deidentified health data associated with over 200,000 admissions to ICUs across the United States between 2014-2015. The database includes vital sign measurements, care plan documentation, severity of illness measures, diagnosis information, and treatment information. Data is collected through the Philips eICU program, a critical care telehealth program that delivers information to caregivers at the bedside.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":5486758129,"compressed_storage_size":5486764761,"resource_type":"Database","access_policy":"Credentialed","topics":["critical care","icu","telemedicine"],"source_url":"https://physionet.org/content/eicu-crd/2.0/"},{"public_project_uuid":"9ef79dfc-f2b9-450b-b83a-ecf9a7e4401b","slug":"nifeadb","version":"1.0.0","core_doi":null,"version_doi":"10.13026/C2CT0S","is_latest_version":true,"short_description":"","publish_date":"2019-02-19","title":"Non-Invasive Fetal ECG Arrhythmia Database","abstract":"Fetal cardiac arrhythmias are defined as any irregular fetal cardiac rhythm or regular rhythm at a rate outside the reference range of 100 to 200 beat per minute (bpm). Arrhythmias are discovered in about 1% of fetuses with about 10% of these being considered potential sources of morbidity. Although most fetal arrhythmias are benign, some can cause fetal hydrops and lead to fetal death. This means that up to 1 fetus in 100 need their arrhythmias to be closely monitored and if indicated treated in-utero using antiarrhythmic therapy.
\n\nThe Non-Invasive Fetal ECG Arrhythmia Database (NIFEA DB) provides a series of fetal arrhythmias recordings (n=12) and a number of control normal rhythm recordings (n=14) performed using the non-invasive fetal electrocardiography (NI-FECG) technique.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":186299569,"compressed_storage_size":186310209,"resource_type":"Database","access_policy":"Open","topics":["fetal","ecg","arrhythmia"],"source_url":"https://physionet.org/content/nifeadb/1.0.0/"},{"public_project_uuid":"ae70fed5-5f67-4225-bcac-2dc33860b33c","slug":"challenge-2000","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2000-02-10","title":"Detecting and Quantifying Apnea Based on the ECG: The PhysioNet/Computing in Cardiology Challenge 2000","abstract":"Obstructive sleep apnea (intermittent cessation of breathing) is a common problem with major health implications, ranging from excessive daytime drowsiness to serious cardiac arrhythmias. Obstructive sleep apnea is associated with increased risks of high blood pressure, myocardial infarction, and stroke, and with increased mortality rates. Standard methods for detecting and quantifying sleep apnea are based on respiration monitoring, which often disturbs or interferes with sleep and is generally expensive. A number of studies during the past 15 years have hinted at the possibility of detecting sleep apnea using features of the electrocardiogram. Such approaches are minimally intrusive, inexpensive, and may be particularly well-suited for screening. The major obstacle to use of such methods is that careful quantitative comparisons of their accuracy against that of conventional techniques for apnea detection have not been published.
\n\nWe therefore offer a challenge to the biomedical research community: demonstrate the efficacy of ECG-based methods for apnea detection using a large, well-characterized, and representative set of data. The goal of the contest is to stimulate effort and advance the state of the art in this clinically significant problem, and to foster both friendly competition and wide-ranging collaborations. We will award prizes of US$500 to the most successful entrant in each of two events.1
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":48123,"compressed_storage_size":52827,"resource_type":"Challenge","access_policy":"Open","topics":["ecg","challenge","apnea"],"source_url":"https://physionet.org/content/challenge-2000/1.0.0/"},{"public_project_uuid":"40495028-d884-49b6-b670-c48c54223568","slug":"challenge-2001","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2001-03-01","title":"Predicting Paroxysmal Atrial Fibrillation/Flutter: The PhysioNet/Computing in Cardiology Challenge 2001","abstract":"Following the success of the first Computers in Cardiology Challenge, we are pleased to offer a new challenge from PhysioNet and Computers in Cardiology 2001. The challenge is to develop a fully automated method to predict the onset of paroxysmal atrial fibrillation/flutter (PAF), based on the ECG prior to the event. The goal of the contest is to stimulate effort and advance the state of the art in this clinically significant problem, and to foster both friendly competition and wide-ranging collaborations.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":50473,"resource_type":"Challenge","access_policy":"Open","topics":["challenge","ecg","atrial fibrillation"],"source_url":"https://physionet.org/content/challenge-2001/1.0.0/"},{"public_project_uuid":"922f4422-f9cd-48f1-b82a-f3aa581ac141","slug":"challenge-2002","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2002-02-14","title":"RR Interval Time Series Modeling: The PhysioNet/Computing in Cardiology Challenge 2002","abstract":"We are pleased to announce the third in our annual series of challenges from PhysioNet and Computers in Cardiology. We received many suggestions for challenge topics, and encourage you to write to us with further suggestions. We chose the topic for this year's challenge not only for its intrinsic interest, but also because it is quite different from the previous two challenges.
\n\nHeart rate variability has attracted much attention from researchers since the early 1980s. It has long been understood that a metronomic heart rate is pathological, and that the healthy heart is influenced by multiple neural and hormonal inputs that result in variations in interbeat (RR) intervals, at time scales ranging from less than a second to 24 hours. Even after 20 years of study, new analytic techniques continue to reveal properties of the time series of RR intervals. Much research in this area aims to discover or to explain how specific changes in variability can be related to specific pathologies.
\n\nGiven how much is known about heart rate variability, it might be thought that simulating a realistic sequence of RR intervals would be an easy task. The intricate interdependencies of variations at different scales, however, make it difficult to create a simulation of sufficient realism to mislead an experienced observer, and it may be even harder to deceive a program designed to quantify these subtle features.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":36807742,"resource_type":"Challenge","access_policy":"Open","topics":["interbeat","challenge","rr interval"],"source_url":"https://physionet.org/content/challenge-2002/1.0.0/"},{"public_project_uuid":"e77ee037-7251-42cd-b060-285819796a2f","slug":"challenge-2003","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2003-03-06","title":"Distinguishing Ischemic from Non-Ischemic ST Changes: The PhysioNet/Computing in Cardiology Challenge 2003","abstract":"For the fourth annual PhysioNet/Computers in Cardiology Challenge, we propose a provocative question of considerable clinical interest: Is it possible to tell the difference between transient ST changes in the ECG that are due to myocardial ischemia, and those that are not?
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":372034,"resource_type":"Challenge","access_policy":"Open","topics":["ischemia","challenge","ecg"],"source_url":"https://physionet.org/content/challenge-2003/1.0.0/"},{"public_project_uuid":"40da732c-1ccd-4028-880c-0499ca1edcfd","slug":"challenge-2004","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2004-10-08","title":"Spontaneous Termination of Atrial Fibrillation: The PhysioNet/Computing in Cardiology Challenge 2004","abstract":"The fifth annual PhysioNet/Computers in Cardiology Challenge focusses on this question: Is it possible to predict if (or when) an episode of atrial fibrillation will end spontaneously?
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":4687005,"resource_type":"Challenge","access_policy":"Open","topics":["challenge","ecg","atrial fibrillation"],"source_url":"https://physionet.org/content/challenge-2004/1.0.0/"},{"public_project_uuid":"cd427961-d361-4589-bfa5-c119f13c0cbc","slug":"challenge-2005","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2005-05-02","title":"The First Five Challenges Revisited: The PhysioNet/Computing in Cardiology Challenge 2005","abstract":"Our popular series of challenges presented in cooperation with Computers in Cardiology continues this year in a new mode. Participants are invited to choose one of the five previous challenges as a topic for a follow-up study:
\n\nThe problems posed by these challenges have continued to interest many researchers long after the final deadlines have passed. We encourage you to read the papers published by previous challenge participants in the Computers in Cardiology proceedings and elsewhere, and to build upon this work in pursuit of even better solutions. This year, we will recognize the most outstanding papers inspired by a previous challenge, with special awards to be presented at Computers in Cardiology.
\n\nWith knowledge gained from previously published papers, can you advance the state of the art further, either by combining efforts with other participants or by entirely new methods?
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":6322,"resource_type":"Challenge","access_policy":"Open","topics":["challenge"],"source_url":"https://physionet.org/content/challenge-2005/1.0.0/"},{"public_project_uuid":"3f17caad-5896-4424-8175-5b1745376362","slug":"challenge-2006","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2006-11-01","title":"QT Interval Measurement: The PhysioNet/Computing in Cardiology Challenge 2006","abstract":"The seventh annual PhysioNet/Computers in Cardiology Challenge addresses a question of high clinical interest: Can the QT interval be measured by fully automated methods with an accuracy acceptable for clinical evaluations?
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":11423623,"resource_type":"Challenge","access_policy":"Open","topics":["challenge","ecg"],"source_url":"https://physionet.org/content/challenge-2006/1.0.0/"},{"public_project_uuid":"f5f5ecf5-14ac-43cc-88c2-d6281684c72d","slug":"challenge-2007","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2007-01-02","title":"Electrocardiographic Imaging of Myocardial Infarction: The PhysioNet/Computing in Cardiology Challenge 2007","abstract":"The aim of Challenge 2007 is to establish how well one can characterize the location and extent of moderate to large, relatively compact infarcts using electrocardiographic evidence (supplemented by a model of the torso geometry and conductivity), in comparison with a "gold standard" expert analysis of gadolinium-enhanced MRI data.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":102249636,"resource_type":"Challenge","access_policy":"Open","topics":["imaging","ecg","challenge"],"source_url":"https://physionet.org/content/challenge-2007/1.0.0/"},{"public_project_uuid":"e932a99d-260f-4f25-a44b-31fba255ba82","slug":"challenge-2008","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2008-03-17","title":"Detecting and Quantifying T-Wave Alternans: The PhysioNet/Computing in Cardiology Challenge 2008","abstract":"The ninth annual PhysioNet/Computers in Cardiology Challenge aims to improve understanding of methods for identification and analysis of T-wave alternans in the ECG.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":85081612,"resource_type":"Challenge","access_policy":"Open","topics":["challenge","ecg"],"source_url":"https://physionet.org/content/challenge-2008/1.0.0/"},{"public_project_uuid":"1a77302a-c9c1-492c-a860-981119855d7a","slug":"challenge-2009","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2009-01-27","title":"Predicting Acute Hypotensive Episodes: The PhysioNet/Computing in Cardiology Challenge 2009","abstract":"This year's challenge is the tenth in the annual series of open challenges hosted by PhysioNet in cooperation with Computers in Cardiology. The goal of the challenge is to predict which patients in the challenge dataset will experience an acute hypotensive episode beginning within the forecast window.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":172457382,"resource_type":"Challenge","access_policy":"Open","topics":["hypertension","challenge","mimic","multiparameter","ehr"],"source_url":"https://physionet.org/content/challenge-2009/1.0.0/"},{"public_project_uuid":"12972ac3-4e4a-4aba-96cc-138dc20ed931","slug":"challenge-2010","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2010-12-01","title":"Mind the Gap: The PhysioNet/Computing in Cardiology Challenge 2010","abstract":"The aim of this year's challenge is to develop robust methods for filling in gaps in multiparameter physiologic data (including ECG signals, continuous blood pressure waveforms, and respiration).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":3282304,"resource_type":"Challenge","access_policy":"Open","topics":["multiparameter","challenge","ecg","missing data"],"source_url":"https://physionet.org/content/challenge-2010/1.0.0/"},{"public_project_uuid":"36021b83-a176-4f74-845c-5040d3af3605","slug":"challenge-2011","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2011-04-19","title":"Improving the Quality of ECGs Collected using Mobile Phones: The PhysioNet/Computing in Cardiology Challenge 2011","abstract":"The aim of the PhysioNet/Computing in Cardiology Challenge 2011 is to develop an efficient algorithm able to run in near real-time within a mobile phone, that can provide useful feedback to a layperson in the process of acquiring a diagnostically useful ECG recording.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":177891351,"resource_type":"Challenge","access_policy":"Open","topics":["cellphone","ecg","challenge"],"source_url":"https://physionet.org/content/challenge-2011/1.0.0/"},{"public_project_uuid":"57823a2d-9065-40fb-8ead-20723e393d7e","slug":"challenge-2012","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2012-01-20","title":"Predicting Mortality of ICU Patients: The PhysioNet/Computing in Cardiology Challenge 2012","abstract":"The focus of the PhysioNet/CinC Challenge 2012 is to develop methods for patient-specific prediction of in-hospital mortality. Participants will use information collected during the first two days of an ICU stay to predict which patients survive their hospitalizations, and which patients do not.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":1972995308,"resource_type":"Challenge","access_policy":"Open","topics":["mortality","prediction","challenge","mimic","ehr"],"source_url":"https://physionet.org/content/challenge-2012/1.0.0/"},{"public_project_uuid":"c4a36fb9-215b-4582-be98-a59fda43ec83","slug":"challenge-2013","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2013-02-21","title":"Noninvasive Fetal ECG: The PhysioNet/Computing in Cardiology Challenge 2013","abstract":"The aim of this year's PhysioNet/Computing in Cardiology Challenge is to encourage development of accurate algorithms for locating QRS complexes and estimating the QT interval in noninvasive FECG signals.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":294460307,"resource_type":"Challenge","access_policy":"Open","topics":["fetal","challenge","ecg"],"source_url":"https://physionet.org/content/challenge-2013/1.0.0/"},{"public_project_uuid":"cc87f936-5138-4744-a41c-a8ad7d327b13","slug":"challenge-2014","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2014-01-07","title":"Robust Detection of Heart Beats in Multimodal Data: The PhysioNet/Computing in Cardiology Challenge 2014","abstract":"This challenge aims to encourage the exploration of robust methods for locating heart beats in continuous long-term data from bedside monitors and similar devices that record not only ECG but usually other physiologic signals as well, including pulsatile signals that directly reflect cardiac activity, and other signals that may have few or no observable markers of heart beats. Our goal is to accelerate development of open-source research tools that can reliably, efficiently, and automatically analyze data such as that contained in the MIMIC II Waveform Database, making use of all relevant information.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":530923917,"resource_type":"Challenge","access_policy":"Open","topics":["multiparameter","challenge","ecg"],"source_url":"https://physionet.org/content/challenge-2014/1.0.0/"},{"public_project_uuid":"e1755e9b-e4ff-4ba5-aa4b-bbebdea278a4","slug":"challenge-2015","version":"1.0.0","core_doi":"10.13026/xxnk-d535","version_doi":"10.13026/c9fg-a467","is_latest_version":true,"short_description":"","publish_date":"2015-02-15","title":"Reducing False Arrhythmia Alarms in the ICU: The PhysioNet/Computing in Cardiology Challenge 2015","abstract":"The 2015 PhysioNet/CinC Challenge aims to encourage the development of algorithms to reduce the incidence of false alarms in the Intensive Care Unit (ICU).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":369541924,"resource_type":"Challenge","access_policy":"Open","topics":["critical care","challenge","icu","arrhythmia","multiparameter","tachycardia","false alarm"],"source_url":"https://physionet.org/content/challenge-2015/1.0.0/"},{"public_project_uuid":"9f1218a5-536a-40ca-bf10-9ffd591227d0","slug":"challenge-2016","version":"1.0.0","core_doi":null,"version_doi":null,"is_latest_version":true,"short_description":"","publish_date":"2016-03-04","title":"Classification of Heart Sound Recordings: The PhysioNet/Computing in Cardiology Challenge 2016","abstract":"The 2016 PhysioNet/CinC Challenge aims to encourage the development of algorithms to classify heart sound recordings collected from a variety of clinical or nonclinical (such as in-home visits) environments. The aim is to identify, from a single short recording (10-60s) from a single precordial location, whether the subject of the recording should be referred on for an expert diagnosis.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1224000245,"compressed_storage_size":1060538323,"resource_type":"Challenge","access_policy":"Open","topics":["sound","challenge","phonocardiogram","heart"],"source_url":"https://physionet.org/content/challenge-2016/1.0.0/"},{"public_project_uuid":"6f5fc234-b785-4cdb-8854-44059505f1fb","slug":"challenge-2017","version":"1.0.0","core_doi":null,"version_doi":"10.13026/d3hm-sf11","is_latest_version":true,"short_description":"","publish_date":"2017-02-01","title":"AF Classification from a Short Single Lead ECG Recording: The PhysioNet/Computing in Cardiology Challenge 2017","abstract":"The 2017 PhysioNet/CinC Challenge aims to encourage the development of algorithms to classify, from a single short ECG lead recording (between 30 s and 60 s in length), whether the recording shows normal sinus rhythm, atrial fibrillation (AF), an alternative rhythm, or is too noisy to be classified.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":0,"compressed_storage_size":1476206149,"resource_type":"Challenge","access_policy":"Open","topics":["ecg","atrial fibrillation","challenge"],"source_url":"https://physionet.org/content/challenge-2017/1.0.0/"},{"public_project_uuid":"0bb4e865-2700-4b30-83ea-a5bb3fbae538","slug":"challenge-2018","version":"1.0.0","core_doi":"10.13026/1q9b-ge17","version_doi":"10.13026/6phb-r450","is_latest_version":true,"short_description":"","publish_date":"2018-02-21","title":"You Snooze You Win: The PhysioNet/Computing in Cardiology Challenge 2018","abstract":"The goal of the challenge is use information from the available signals to correctly classify target arousal regions.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":286233571088,"compressed_storage_size":48042063,"resource_type":"Challenge","access_policy":"Open","topics":["sleep","challenge","polysomnography","circadian","apnea"],"source_url":"https://physionet.org/content/challenge-2018/1.0.0/"},{"public_project_uuid":"39d4f31c-dee5-466c-a74f-b41a09a5a9de","slug":"mednli-bionlp19","version":"1.0.0","core_doi":"10.13026/2mx7-kt26","version_doi":"10.13026/9pk3-y617","is_latest_version":false,"short_description":"MedNLI test set for Shared Task at BioNLP 2019","publish_date":"2019-04-17","title":"MedNLI for Shared Task at ACL BioNLP 2019","abstract":"Natural language inference (NLI) is the task of determining whether a given hypothesis can be inferred from a given premise. Also known as Recognizing Textual Entailment (RTE), this task has enjoyed popularity among researchers for a long time. However, almost all datasets for this task focused on open domain data such as as news texts, blogs, and so on. To address this gap, the MedNLI dataset was created or language inference in the medical domain. MedNLI was a derived dataset with data sourced from MIMIC-III v1.4. In order to stimulate research for this problem, the MEDIQA shared task has been organized at BioNLP 2019. The dataset provided herein is a test set of 405 premise hypothesis pairs for the NLI challenge at the MEDIQA shared task. Participants of the shared task are expected to use the MedNLI data for development of their models and this dataset will be used as an unseen dataset for scoring each participant submission.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":370404,"compressed_storage_size":42140,"resource_type":"Database","access_policy":"Credentialed","topics":["natural language inference","recognizing textual entailment","mimic"],"source_url":"https://physionet.org/content/mednli-bionlp19/1.0.0/"},{"public_project_uuid":"db96cfcb-ef40-4680-9ab2-267be1cbcf05","slug":"mimiciii-demo","version":"1.4","core_doi":"10.13026/jbmn-w042","version_doi":"10.13026/C2HM2Q","is_latest_version":true,"short_description":"An open source demo of the MIMIC-III Clinical Database","publish_date":"2019-04-24","title":"MIMIC-III Clinical Database Demo","abstract":"MIMIC-III is a large, freely-available database comprising deidentified health-related data associated with over 40,000 patients who stayed in critical care units of the Beth Israel Deaconess Medical Center between 2001 and 2012 [1]. The MIMIC-III Clinical Database is available on PhysioNet (doi: 10.13026/C2XW26). Though deidentified, MIMIC-III contains detailed information regarding the care of real patients, and as such requires credentialing before access. To allow researchers to ascertain whether the database is suitable for their work, we have manually curated a demo subset, which contains information for 100 patients also present in the MIMIC-III Clinical Database. Notably, the demo dataset does not include free-text notes.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":107953726,"compressed_storage_size":14093220,"resource_type":"Database","access_policy":"Open","topics":["electronic health records","mimic","critical care"],"source_url":"https://physionet.org/content/mimiciii-demo/1.4/"},{"public_project_uuid":"77d56415-33fd-4708-8d8d-83f530d81424","slug":"bhi-2018-challenge","version":"1.0","core_doi":"10.13026/5bhj-nf42","version_doi":"10.13026/v1jk-ax96","is_latest_version":true,"short_description":"A challenge to explore real clinical questions in critically ill patients using the MIMIC. A collaboration with the IEEE Conference on Biomedical and Health Informatics 2018 and the IEEE Conference on Body Sensor Networks.","publish_date":"2018-02-05","title":"2018 IEEE BHI and BSN Data Challenge","abstract":"In collaboration with the IEEE Conference on Biomedical and Health Informatics (BHI) 2018 and the IEEE Conference on Body Sensor Networks (BSN), we are hosting a challenge to explore real clinical questions in critically ill patients using the MIMIC-III database. Participants in the challenge will be invited to present at the BHI & BSN Annual Conference in Las Vegas, USA (4-7 March 2018): https://bhi-bsn.embs.org/2018/
","license":{"name":"MIT License"},"dua":null,"main_storage_size":5702856,"compressed_storage_size":5704998,"resource_type":"Challenge","access_policy":"Open","topics":["challenge","weekend effect","ieee","mimic"],"source_url":"https://physionet.org/content/bhi-2018-challenge/1.0/"},{"public_project_uuid":"3fd4543a-c99f-4419-a226-333d3cd1fbf3","slug":"eicu-crd-demo","version":"2.0","core_doi":"10.13026/krfm-4167","version_doi":"10.13026/gxmm-es70","is_latest_version":false,"short_description":"An openly available subset of the eICU Collaborative Research Database.","publish_date":"2019-07-15","title":"eICU Collaborative Research Database Demo","abstract":"This project is a demo version of the eICU Collaborative Research Database, a multi-center database comprised of deidentified health data for over 200,000 admissions to ICUs across the United States between 2014-2015. The database includes vital sign measurements, care plan documentation, severity of illness measures, diagnosis information, and treatment information. Data is collected through the Philips eICU program, a critical care telehealth program that delivers information to caregivers at the bedside. This open access demo allows researchers to ascertain whether the eICU Collaborative Research Database is suitable for their work. It includes over 2,500 unit stays selected from 20 of the larger hospitals in the eICU Collaborative Research Database.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":54780340,"compressed_storage_size":54728526,"resource_type":"Database","access_policy":"Open","topics":["eicu","open data","critical care","medical informatics"],"source_url":"https://physionet.org/content/eicu-crd-demo/2.0/"},{"public_project_uuid":"ef256b61-a2e4-44d7-9e43-1b750b1a40bc","slug":"hss","version":"1.0","core_doi":"10.13026/9jqc-rv39","version_doi":"10.13026/vnt9-kf93","is_latest_version":true,"short_description":"Heart sound segmentation code, based on a duration-dependent hidden Markov model, extended with the use of logistic regression for emission probability estimation and an enhanced Viterbi algorithm.","publish_date":"2019-07-29","title":"Logistic Regression-HSMM-based Heart Sound Segmentation","abstract":"The identification of the exact positions of the first and second heart sounds within a phonocardiogram (PCG), or heart sound segmentation, is an essential step in the automatic analysis of heart sound recordings, allowing for the classification of pathological events. While threshold-based segmentation methods have shown modest success, probabilistic models, such as hidden Markov models, have recently been shown to surpass the capabilities of previous methods. Segmentation performance is further improved when apriori information about the expected duration of the states is incorporated into the model, such as in a hidden semi-Markov model (HSMM). This code provides Matlab algorithms to perform HSMM-based heart sound segmentation.
","license":{"name":"GNU General Public License version 3"},"dua":null,"main_storage_size":155100764,"compressed_storage_size":154922123,"resource_type":"Software","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/hss/1.0/"},{"public_project_uuid":"14b4578d-4049-43ec-ae5d-35458924a7ae","slug":"ct-ich","version":"1.0.0","core_doi":"10.13026/g1kq-sc84","version_doi":"10.13026/w8q8-ky94","is_latest_version":false,"short_description":"Head computed tomography scans with intracranial hemmorhage segmentation.","publish_date":"2019-08-01","title":"Computed Tomography Images for Intracranial Hemorrhage Detection and Segmentation","abstract":"After traumatic brain injury (TBI), intracranial hemorrhage (ICH) may occur that could lead to death or disability if it is not accurately diagnosed and treated in a time-sensitive procedure. Currently, Computerized Tomography (CT) scans are examined by radiologists to diagnose ICH and localize affected regions. In this work, we collected a dataset of 82 CT scans of patients with TBI. ICH regions in these scans were delineated in each slice by two radiologists. The radiologists also annotated each CT slice for the presence of different types of ICH and fractures.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":107586941,"compressed_storage_size":93964224,"resource_type":"Database","access_policy":"Open","topics":["jpg","segmentation","computed tomography","intracranial hemmorhage"],"source_url":"https://physionet.org/content/ct-ich/1.0.0/"},{"public_project_uuid":"3103bc95-c282-401d-a34a-f7cc6949527b","slug":"mimic-cxr","version":"1.0.0","core_doi":"10.13026/s5dg-6s42","version_doi":"10.13026/C2JT1Q","is_latest_version":false,"short_description":"Chest radiographs in JPG format with structured labels extracted from the free-text reports.","publish_date":"2019-08-02","title":"MIMIC-CXR Database","abstract":"The MIMIC Chest X-ray (MIMIC-CXR) Database v1.0.0 is a large publicly available dataset of chest radiographs with structured labels. The dataset contains 371,920 images corresponding to 224,548 radiographic studies performed at the Beth Israel Deaconess Medical Center in Boston, MA. The dataset is de-identified to satisfy the US Health Insurance Portability and Accountability Act of 1996 (HIPAA) Safe Harbor requirements. Protected health information (PHI) has been removed. The dataset is intended to support a wide body of research in medicine including image understanding, natural language processing, and decision support.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":2595,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":["machine learning","mimic","chest x-rays","radiology","natural language processing","computer vision"],"source_url":"https://physionet.org/content/mimic-cxr/1.0.0/"},{"public_project_uuid":"52cda938-92f5-4a03-acda-adc0d3b6a227","slug":"challenge-2019","version":"1.0.0","core_doi":"10.13026/0y7p-s471","version_doi":"10.13026/v64v-d857","is_latest_version":true,"short_description":"The 2019 PhysioNet Computing in Cardiology Challenge invites participants to predict sepsis in clinical data","publish_date":"2019-08-05","title":"Early Prediction of Sepsis from Clinical Data: The PhysioNet/Computing in Cardiology Challenge 2019","abstract":"Sepsis is a life-threatening condition that occurs when the body's response to infection causes tissue damage, organ failure, or death (Singer et al., 2016). In the U.S., nearly 1.7 million people develop sepsis and 270,000 people die from sepsis each year; over one third of people who die in U.S. hospitals have sepsis (CDC). Internationally, an estimated 30 million people develop sepsis and 6 million people die from sepsis each year; an estimated 4.2 million newborns and children are affected (WHO). Sepsis costs U.S. hospitals more than any other health condition at $24 billion (13% of U.S. healthcare expenses) a year, and a majority of these costs are for sepsis patients that were not diagnosed at admission (Paoli et al., 2018). Sepsis costs are even greater globally with the developing world at most risk. Altogether, sepsis is a major public health issue responsible for significant morbidity, mortality, and healthcare expenses.
\n\nEarly detection and antibiotic treatment of sepsis are critical for improving sepsis outcomes, where each hour of delayed treatment has been associated with roughly an 4-8% increase in mortality (Kumar et al., 2006; Seymour et al., 2017). To help address this problem, clinicians have proposed new definitions for sepsis (Singer et al., 2016), but the fundamental need to detect and treat sepsis early still remains, and basic questions about the limits of early detection remain unanswered. The PhysioNet/Computing in Cardiology Challenge 2019 provides an opportunity to address these questions.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":335112,"compressed_storage_size":0,"resource_type":"Challenge","access_policy":"Open","topics":["challenge","sepsis","prediction"],"source_url":"https://physionet.org/content/challenge-2019/1.0.0/"},{"public_project_uuid":"c5847384-7d14-41df-a0c0-dfb3f1ae81ac","slug":"deidentifiedmedicaltext","version":"1.0","core_doi":"10.13026/rma9-zb29","version_doi":"10.13026/jc2a-ca12","is_latest_version":true,"short_description":"Gold standard corpus of 2,434 deidentified nursing notes","publish_date":"2007-12-18","title":"Deidentified Medical Text","abstract":"Available here is a gold standard corpus of 2,434 nursing notes that have been thoroughly deidentified by a multi-pass process that included meticulous reviews by three or more experts working independently, as well as by a variety of automated methods. All detected instances of PHI in these nursing notes have been replaced by realistic surrogate data. The gold standard corpus is currently available only to those who have been granted access to PhysioNet Clinical Databases.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":4342079,"compressed_storage_size":1688870,"resource_type":"Database","access_policy":"Credentialed","topics":["medical text","nursing notes","de-identification","hipaa"],"source_url":"https://physionet.org/content/deidentifiedmedicaltext/1.0/"},{"public_project_uuid":"f9044df7-7441-4657-b772-e4759cb87d43","slug":"physiozoo","version":"1.0.0","core_doi":"10.13026/p19w-dp86","version_doi":"10.13026/p63q-hq95","is_latest_version":true,"short_description":"PhysioZoo is a collaborative platform dedicated to the study of the heart rate variability in electrophysiological recordings from mammals","publish_date":"2019-08-27","title":"PhysioZoo - mammalian NSR databases","abstract":"The PhysioZoo database contains electrocardiographic recordings (ECG) taken from multiple types of mammals (dogs, rabbits, mice). Each record is provided in 3 different formats: MATLAB (.mat), simple text (.txt), PhysioNet (.hea,.dat,.qrs). For each record the following data is available: the raw ECG signal, reference (human corrected) R-peak annotations and the signal quality annotations. All of the files formats are compatible with the PhysioZoo open source software available at physiozoo.com. When the recordings were taken, all the mammals were conscious, and no drugs were administered prior to the recordings.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":420809156,"compressed_storage_size":123005476,"resource_type":"Database","access_policy":"Open","topics":["heart rate variabillity","ecg","electrophysiology","mammals"],"source_url":"https://physionet.org/content/physiozoo/1.0.0/"},{"public_project_uuid":"98885acf-0b67-45a0-a6b4-a310231507cd","slug":"mimic-cxr","version":"2.0.0","core_doi":"10.13026/s5dg-6s42","version_doi":"10.13026/C2JT1Q","is_latest_version":false,"short_description":"Chest radiographs in DICOM format with associated free-text reports.","publish_date":"2019-09-19","title":"MIMIC-CXR Database","abstract":"The MIMIC Chest X-ray (MIMIC-CXR) Database v2.0.0 is a large publicly available dataset of chest radiographs in DICOM format with free-text radiology reports. The dataset contains 377,110 images corresponding to 227,835 radiographic studies performed at the Beth Israel Deaconess Medical Center in Boston, MA. The dataset is de-identified to satisfy the US Health Insurance Portability and Accountability Act of 1996 (HIPAA) Safe Harbor requirements. Protected health information (PHI) has been removed. The dataset is intended to support a wide body of research in medicine including image understanding, natural language processing, and decision support.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":5011060220000,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":["machine learning","natural language processing","computer vision","mimic","radiology","chest x-rays"],"source_url":"https://physionet.org/content/mimic-cxr/2.0.0/"},{"public_project_uuid":"fcc37b2d-c1eb-4594-87d8-f2759ee95641","slug":"mednli","version":"1.0.0","core_doi":"10.13026/c6kh-4h33","version_doi":"10.13026/C2RS98","is_latest_version":true,"short_description":"This is a resource for training machine learning models for language inference in the medical domain.","publish_date":"2019-10-01","title":"MedNLI - A Natural Language Inference Dataset For The Clinical Domain","abstract":"State of the art models using deep neural networks have become very good in learning an accurate mapping from inputs to outputs. However, they still lack generalization capabilities in conditions that differ from the ones encountered during training. This is even more challenging in specialized, and knowledge intensive domains, where training data is limited. To address this gap, we introduce MedNLI - a dataset annotated by doctors, performing a natural language inference task (NLI), grounded in the medical history of patients. As the source of premise sentences, we used the MIMIC-III. More specifically, to minimize the risks to patient privacy, we worked with clinical notes corresponding to the deceased patients. The clinicians in our team suggested the Past Medical History to be the most informative section of a clinical note, from which useful inferences can be drawn about the patient.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":13817812,"compressed_storage_size":1516111,"resource_type":"Database","access_policy":"Credentialed","topics":["natural language inference","recognizing textual entailment"],"source_url":"https://physionet.org/content/mednli/1.0.0/"},{"public_project_uuid":"df254d31-36a0-4893-863e-bea6377874d5","slug":"sleep-accel","version":"1.0.0","core_doi":"10.13026/50yh-tt09","version_doi":"10.13026/hmhs-py35","is_latest_version":true,"short_description":"Motion data and heart rate measurements from Apple Watches from sleeping people undergoing polysomnography.","publish_date":"2019-10-08","title":"Motion and heart rate from a wrist-worn wearable and labeled sleep from polysomnography","abstract":"This project contains acceleration (in units of g) and heart rate (bpm, measured from photoplethysmography) recorded from the Apple Watch, as well as labeled sleep scored from gold-standard polysomnography. Data were collected at the University of Michigan from June 2017 to March 2019, and there are 31 subjects in total. Code to read and process these files is available on GitHub. The paper corresponding to the work is Walch et al., "Sleep stage prediction with raw acceleration and photoplethysmography heart rate data derived from a consumer wearable device", SLEEP (2019).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":2348136828,"compressed_storage_size":576850847,"resource_type":"Database","access_policy":"Open","topics":["polysomnography","sleep"],"source_url":"https://physionet.org/content/sleep-accel/1.0.0/"},{"public_project_uuid":"916236b6-891f-4d63-8bf9-10258d3b306a","slug":"phdsm","version":"1.1","core_doi":"10.13026/5889-jj86","version_doi":"10.130264/c23h3b","is_latest_version":true,"short_description":"Conductivity and relative permittivity of healthy and diseased skeletal muscle.","publish_date":"2019-11-12","title":"Permittivity of Healthy and Diseased Skeletal Muscle","abstract":"A better understanding of the permittivity of skeletal muscle is essential for the development of new diagnostic tools and approaches for neuromuscular evaluation. However, there remain important knowledge gaps in our understanding of this property in healthy and diseased skeletal muscle, which hinder its translation into clinical application. The aim of this project is to provide a platform for the researchers to contribute and share the dissemination of this property in healthy and diseased muscle. Ultimately, the normative data reported will offer the scientific community the opportunity to improve the accuracy of existing techniques as well as developing new diagnostic tools and therapies.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":99771,"compressed_storage_size":50437,"resource_type":"Database","access_policy":"Open","topics":["anisotropy","neuromuscular disorders","skeletal muscle","permittivity"],"source_url":"https://physionet.org/content/phdsm/1.1/"},{"public_project_uuid":"5dbecc7e-8046-4862-bf20-eaa9427d0db1","slug":"wctecgdb","version":"1.0.1","core_doi":"10.13026/2mz7-3v18","version_doi":"10.13026/f73z-an96","is_latest_version":true,"short_description":"Wilson Central Terminal ECG signals recorded from 92 patients.","publish_date":"2019-11-13","title":"Wilson Central Terminal ECG Database","abstract":"The Wilson Central Terminal ECG Database (WCTECGdb) contains the Wilson Central Terminal (WCT), the voltage of right arm (RA), left arm (LA) and left leg (LL), six unipolar chest leads associated with three limb leads and six precordial leads.
\n\nWith our ECG device, the potentials of RA, LA, LL and six true unipolar leads are measured using the right leg as the reference point. Having this configuration, we have been able to overcome the difficulty of recording the WCT in a clinical setting. Our data confirmed once again that the WCT does not have a small value, and could be as high as 241% of lead II and has standard ECG characteristics such as the p-wave and the t-wave. The dataset comprises of 540 ten-second segments recorded from 92 patients.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":321041053,"compressed_storage_size":317268304,"resource_type":"Database","access_policy":"Open","topics":["ecg","electrocardiography","wilson central terminal","limb potential","unipolar lead"],"source_url":"https://physionet.org/content/wctecgdb/1.0.1/"},{"public_project_uuid":"de93f1f5-1f63-419d-923b-5931c3725c6b","slug":"mimic-cxr-jpg","version":"2.0.0","core_doi":"10.13026/th9c-ae10","version_doi":"10.13026/8360-t248","is_latest_version":false,"short_description":"Chest x-rays in JPG format with structured labels derived from the associated radiology report.","publish_date":"2019-11-14","title":"MIMIC-CXR-JPG - chest radiographs with structured labels","abstract":"The MIMIC Chest X-ray JPG (MIMIC-CXR-JPG) Database v2.0.0 is a large publicly available dataset of chest radiographs in JPG format with structured labels derived from free-text radiology reports. The MIMIC-CXR-JPG dataset is wholly derived from MIMIC-CXR, providing JPG format files derived from the DICOM images and structured labels derived from the free-text reports. The aim of MIMIC-CXR-JPG is to provide a convenient processed version of MIMIC-CXR, as well as to provide a standard reference for data splits and image labels. The dataset contains 377,110 JPG format images and structured labels derived from the 227,827 free-text radiology reports associated with these images. The dataset is de-identified to satisfy the US Health Insurance Portability and Accountability Act of 1996 (HIPAA) Safe Harbor requirements. Protected health information (PHI) has been removed. The dataset is intended to support a wide body of research in medicine including image understanding, natural language processing, and decision support.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":598683852000,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":["mimic","computer vision","radiology","deep learning","chest x-ray"],"source_url":"https://physionet.org/content/mimic-cxr-jpg/2.0.0/"},{"public_project_uuid":"83a2fe9a-8837-40d1-87f4-a02a4227036a","slug":"ct-ich","version":"1.2.0","core_doi":"10.13026/g1kq-sc84","version_doi":"10.13026/w8q8-ky94","is_latest_version":false,"short_description":"Head computed tomography (CT) scans with intracranial hemorrhage (ICH) segmentation, ICH subtypes and skull fracture.","publish_date":"2019-11-14","title":"Computed Tomography Images for Intracranial Hemorrhage Detection and Segmentation","abstract":"After traumatic brain injury (TBI), intracranial hemorrhage (ICH) may occur that could lead to death or disability if it is not accurately diagnosed and treated in a time-sensitive procedure. Currently, Computerized Tomography (CT) scans are examined by radiologists to diagnose intracranial hemorrhage to localize affected regions. In this work, we collected a dataset of 82 CT scans of patients with traumatic brain injury. Intracranial hemorrhage regions in these scans were delineated in each slice by two radiologists. The radiologists also annotated each CT slice for the presence of different types of intracranial hemorrhage and fracture.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":0,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Open","topics":["ich","skull fracture","computed tomography","intracranial hemmorhage","segmentation","nifti"],"source_url":"https://physionet.org/content/ct-ich/1.2.0/"},{"public_project_uuid":"10920975-a802-48b1-86d8-69a0df61efab","slug":"chaos-heart-rate","version":"1.0.0","core_doi":"10.13026/hvnd-ex27","version_doi":"10.13026/6gb6-8446","is_latest_version":true,"short_description":"In its June 2008 issue, the editors of Chaos announced a new feature, \"Controversial Topics in Nonlinear Dynamics\". The first controversial topic to be aired is \"Is the Normal Heart Rate Chaotic?\". This project provides data to explore this question.","publish_date":"2008-10-30","title":"Is the normal heart rate chaotic?","abstract":"In its June 2008 issue, the editors of Chaos announced a new feature, "Controversial Topics in Nonlinear Dynamics". The first controversial topic to be aired is "Is the Normal Heart Rate Chaotic?". Since major impediments to achieving consensus on the statistical properties of biologic time series have included the lack of open access time series, PhysioNet has provided a set of 15 heart beat (RR-interval) time series in health and disease (congestive heart failure and atrial fibrillation). Each time series is about 24 hours long (roughly 100,000 intervals), and is provided as a text file for analysis and as an image for visual review. All of the time series were derived from continuous ambulatory (Holter) electrocardiograms. We encourage participants in the Chaos Controversies issue to use the 15 RR interval time series provided here as a common focus of analysis.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":58828322,"compressed_storage_size":20673042,"resource_type":"Challenge","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/chaos-heart-rate/1.0.0/"},{"public_project_uuid":"98834464-259b-4636-a8e2-04fad1540f78","slug":"mednli-bionlp19","version":"1.0.1","core_doi":"10.13026/2mx7-kt26","version_doi":"10.13026/gtv4-g455","is_latest_version":true,"short_description":"Data for the MedNLI Shared Task at the 2019 ACL BioNLP 2019 Workshop on Biomedical Language Processing","publish_date":"2019-11-29","title":"MedNLI for Shared Task at ACL BioNLP 2019","abstract":"Natural Language Inference (NLI) is the task of determining whether a given hypothesis can be inferred from a given premise. Also known as Recognizing Textual Entailment (RTE), this task has enjoyed popularity among researchers for some time. However, almost all datasets for this task focused on open domain data such as as news texts, blogs, and so on. To address this gap, the MedNLI dataset was created for language inference in the medical domain. MedNLI is a derived dataset with data sourced from MIMIC-III v1.4. In order to stimulate research for this problem, a shared task on Medical Inference and Question Answering (MEDIQA) was organized at the workshop for biomedical natural language processing (BioNLP) 2019. The dataset provided herein is a test set of 405 premise hypothesis pairs for the NLI challenge in the MEDIQA shared task. Participants of the shared task are expected to use the MedNLI data for development of their models and this dataset was used as an unseen dataset for scoring each participant submission.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":392999,"compressed_storage_size":45598,"resource_type":"Database","access_policy":"Credentialed","topics":["recognizing textual entailment","natural language inference","mimic"],"source_url":"https://physionet.org/content/mednli-bionlp19/1.0.1/"},{"public_project_uuid":"388bc02c-4004-459e-8af4-e7ebc24e4ec2","slug":"picdb","version":"1.0.0","core_doi":"10.13026/c9gh-n952","version_doi":"10.13026/bcbd-4t11","is_latest_version":false,"short_description":"PIC (Paediatric Intensive Care) is a large paediatric-specific, single-centre, bilingual database comprising information relating to children admitted to critical care units at a large children’s hospital in China.","publish_date":"2019-12-01","title":"Paediatric Intensive Care database","abstract":"The Paediatric Intensive Care (PIC) database is a large paediatric-specific, single-center, bilingual database comprising information relating to children admitted to critical care units at a large children’s hospital in China. The database is de-identified and includes vital sign measurements, medications, laboratory measurements, fluid balance recordings, diagnostic codes, demographic information, and more. The data are publicly available after credentialing, which includes completion of a training course on research with human subjects and signing of a data use agreement mandating responsible handling of the data and adherence to the principle of collaborative research. The PIC database builds upon the success of the widely used MIMIC (Medical Information Mart for Intensive Care) database, and extends the approach into the field of paediatric critical care. The database has many unique characteristics which support academic and industrial research including the development of machine learning algorithms, clinical decision support tools, quality improvement initiatives, and international data sharing.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":192695462,"compressed_storage_size":191616134,"resource_type":"Database","access_policy":"Credentialed","topics":["pediatrics","critical care","natural language processing","intensive care"],"source_url":"https://physionet.org/content/picdb/1.0.0/"},{"public_project_uuid":"0e0851ff-7a8c-4ded-bed2-c41c00b45ce4","slug":"ct-ich","version":"1.3.0","core_doi":"10.13026/g1kq-sc84","version_doi":"10.13026/w8q8-ky94","is_latest_version":false,"short_description":"Head computed tomography (CT) scans with intracranial hemorrhage (ICH) segmentation, ICH subtypes and skull fracture.","publish_date":"2019-12-20","title":"Computed Tomography Images for Intracranial Hemorrhage Detection and Segmentation","abstract":"After traumatic brain injury (TBI), intracranial hemorrhage (ICH) may occur that could lead to death or disability if it is not accurately diagnosed and treated in a time-sensitive procedure. Currently, Computerized Tomography (CT) scans are examined by radiologists to diagnose intracranial hemorrhage to localize affected regions. In this work, we collected a dataset of 82 CT scans of patients with traumatic brain injury. Intracranial hemorrhage regions in these scans were delineated in each slice by two radiologists. The radiologists also annotated each CT slice for the presence of different types of intracranial hemorrhage and fracture.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":0,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Open","topics":["ich","segmentation","intracranial hemmorhage","nifti","computed tomography","skull fracture"],"source_url":"https://physionet.org/content/ct-ich/1.3.0/"},{"public_project_uuid":"13358792-f3f8-48e7-befd-7cf4d348385f","slug":"shareclefehealth2013","version":"1.0","core_doi":"10.13026/rxa7-q798","version_doi":"10.13026/0zsp-0e97","is_latest_version":true,"short_description":"2013 ShARe/CLEF eHealth Evaluation Lab: Natural Language Processing and Information Retrieval for Clinical Care (Tasks 1 and 2).","publish_date":"2013-02-15","title":"ShAReCLEF eHealth 2013: Natural Language Processing and Information Retrieval for Clinical Care","abstract":"This is the pilot year of the ShARe/CLEF eHealth Evaluation Lab, a shared task focused on natural language processing (NLP) and information retrieval (IR) for clinical care. The task is co-organized by the Shared Annotated Resources (ShARe) project and the CLEF Initiative (Conference and Labs of the Evaluation Forum, formerly known as Cross-Language Evaluation Forum). The vision of ShARe/CLEF is two-fold: (1) to develop tasks that potentially impact patient understanding of medical information and (2) to provide the community with an increasingly sophisticated dataset of clinical narrative to advance the state-of-the-art in Natural Language Processing, Information Extraction and Information Retrieval in healthcare.
\n\nThe ShARe/CLEF eHealth 2013 Challenge has three tasks: Task 1 and Task 2 involve annotation of entities in a set of narrative clinical reports; Task 3 involves retrieval of web pages based on queries generated when reading the clinical reports. The datasets for Tasks 1 and 2 comprise de-identified clinical free-text notes from the MIMIC-II database (version 2.5). The dataset for Task 3 consists of a set of medical-related documents provided by the Khresmoi E.U. Consortium. This project provides data files for Task 1 and Task 2 only.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":9111759,"compressed_storage_size":8504830,"resource_type":"Challenge","access_policy":"Credentialed","topics":["natural language processing"],"source_url":"https://physionet.org/content/shareclefehealth2013/1.0/"},{"public_project_uuid":"6abaa90a-12b5-4022-aa16-4286b660f36c","slug":"shareclefehealth2014task2","version":"1.0","core_doi":"10.13026/nv8n-1164","version_doi":"10.13026/0zgk-9j94","is_latest_version":true,"short_description":"The ShARe/CLEF eHealth 2014 Challenge (Task 2) on Disorder Attributes in Clinical Reports","publish_date":"2013-11-01","title":"ShAReCLEF eHealth Evaluation Lab 2014 (Task 2): Disorder Attributes in Clinical Reports","abstract":"This project focuses on Task 2 of the 2014 ShARe/CLEF eHealth evaluation lab, which extended Task 1 of the 2013 ShARe/CLEF eHealth evaluation lab by focusing on identification and normalization of disorder attributes. The task comprises two subtasks: attribute normalization (task 2a) and cue identification (task 2b). We instructed participants to develop a system which either kept or updated a default attribute value for each task. Participant systems were evaluated against a blind reference standard of 133 discharge summaries using Accuracy (task 2a) and F-score (task 2b).
\n\nThis is the second year of the ShARe/CLEF eHealth Evaluation Lab, a shared task focused on natural language processing (NLP) and information retrieval (IR) for clinical care. The task is co-organized by the Shared Annotated Resources (ShARe) project and the CLEF Initiative (Conference and Labs of the Evaluation Forum, formerly known as Cross-Language Evaluation Forum). The vision of ShARe/CLEF is two-fold: (1) to develop tasks that potentially impact patient understanding of medical information and (2) to provide the community with an increasingly sophisticated dataset of clinical narrative to advance the state-of-the-art in Natural Language Processing, Information Extraction and Information Retrieval in healthcare.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":82891278,"compressed_storage_size":82187532,"resource_type":"Challenge","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/shareclefehealth2014task2/1.0/"},{"public_project_uuid":"7066620d-1839-4e8e-af8e-0ec41edb1780","slug":"widsdatathon2020","version":"1.0.0","core_doi":"10.13026/530d-1e61","version_doi":"10.13026/vc0e-th79","is_latest_version":true,"short_description":"WiDS (Women in Data Science) Datathon 2020: ICU Mortality Prediction focuses on patient health. Join a team, explore the data, and share your insights: http://bit.ly/WiDSdatathon2020","publish_date":"2020-01-22","title":"WiDS (Women in Data Science) Datathon 2020: ICU Mortality Prediction","abstract":"WiDS (Women in Data Science) Datathon 2020: ICU Mortality Prediction focuses on patient health through data from MIT’s GOSSIS (Global Open Source Severity of Illness Score) initiative. Brought to you by the Global WiDS team, the West Big Data Innovation Hub, and the WiDS Datathon Committee, this year’s datathon is launching on Kaggle: bit.ly/WiDSdatathon2020kaggle.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":336009,"compressed_storage_size":0,"resource_type":"Challenge","access_policy":"Restricted","topics":["mortality","data science","kaggle","predictive analytics","risk","women in data science","challenge","icu"],"source_url":"https://physionet.org/content/widsdatathon2020/1.0.0/"},{"public_project_uuid":"8af6820f-8165-4073-9650-dcd8dc348684","slug":"cerebral-vasoreg-diabetes","version":"1.0.0","core_doi":"10.13026/c8ax-aq77","version_doi":"10.13026/m40k-4758","is_latest_version":true,"short_description":"Diabetes is a risk factor for cerebral hypoperfustion and microvascular disease. This study assessed the effects of diabetic autonomic neuropathy with OH on cerebral vasoregulation.","publish_date":"2020-01-30","title":"Cerebral Vasoregulation in Diabetes","abstract":"This observational study evaluated the effects of type 2 diabetes on cerebral vasoregulation and functional outcomes, measured by blood flow responses to hypocapnia and hypercapnia, Valsalva maneuver, head-up tilt, and sit-to-stand test. This dataset contains 37 diabetic participants and 49 controls (aged 55 to 75 years) with continuous measurements of cerebral blood flow using transcranial Doppler and MRI, heart rate, blood pressure, and respiratory parameters, balance, walking, laboratory and retinopathy measures.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":3265620876,"compressed_storage_size":2138762662,"resource_type":"Database","access_policy":"Open","topics":["type 2 diabetes","perfusion-arterial spin labeling mri","cerebral vasoregulation","white matter hyperintensities","orthostatic hypotension","cerebral blood flow- transcranial doppler"],"source_url":"https://physionet.org/content/cerebral-vasoreg-diabetes/1.0.0/"},{"public_project_uuid":"066cfb33-f4ec-49ac-ae63-af93729d7a8d","slug":"challenge-2020","version":"1.0.0","core_doi":"10.13026/m77n-sx13","version_doi":"10.13026/mpha-2q23","is_latest_version":false,"short_description":"The goal of the 2020 PhysioNet - Computing in Cardiology Challenge is to design and implement a working, open-source algorithm that can automatically identify cardiac abnormalities in 12-lead ECG recordings.","publish_date":"2020-02-16","title":"Classification of 12-lead ECGs: The PhysioNet/Computing in Cardiology Challenge 2020","abstract":"The standard 12-lead ECG has been widely used to diagnose a variety of cardiac abnormalities such as cardiac arrhythmias and predicts cardiovascular morbidity and mortality. The early and correct diagnosis of cardiac abnormalities can increase the chances of successful treatments. However, manual interpretation of the electrocardiogram is time-consuming and requires skilled personnel with a high degree of training. Automatic detection and classification of cardiac abnormalities can assist physicians in the diagnosis of the growing number of ECGs recorded. The PhysioNet/Computing in Cardiology Challenge 2020 provides an opportunity to address this problem by providing data from a wide set of sources.
\n\nPlease see https://physionetchallenges.github.io/2020/ for all information about this year's Challenge. We are using the above GitHub Pages link and Google Groups to post all updates this year. At the end of the Challenge, the current page will be updated to reflect the complete event and the final results.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":14919,"compressed_storage_size":0,"resource_type":"Challenge","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/challenge-2020/1.0.0/"},{"public_project_uuid":"b78ea6c9-4ae3-49e0-bc6c-e831ffe1f251","slug":"phenotype-annotations-mimic","version":"1.20.3","core_doi":"10.13026/q0wp-m823","version_doi":"10.13026/txmt-8m40","is_latest_version":true,"short_description":"Clinical notes, annotated by at least two expert annotators for over ten patient phenotypes, including advanced cancer, substance abuse, and treatment non-adherence.","publish_date":"2020-03-05","title":"Phenotype Annotations for Patient Notes in the MIMIC-III Database","abstract":"A crucial step within secondary analysis of electronic health records (EHRs) is to identify the patient cohort under investigation. While EHRs contain medical billing codes that aim to represent the conditions and treatments patients may have, much of the information is only present in the patient notes. Therefore, it is critical to develop robust algorithms to infer patients' conditions and treatments from their written notes.
\n\nWe introduce a dataset for patient phenotyping, a task that is defined as the identification whether a patient has a given phenotype (also referred to as indication) based on their patient note. Patient notes of MIMIC-III, a dataset collected from Intensive Care Units of a large tertiary care hospital in Boston, were manually annotated for the presence of several high-context phenotypes relevant to treatment and risk of re-hospitalization.
\n\nEach note has been annotated by two expert human annotators (one clinical researcher and one resident physician). Annotated phenotypes include treatment non-adherence, chronic pain, advanced/metastatic cancer, as well as 10 other phenotypes. This dataset can be utilized for academic and industrial research in medicine and computer science, particularly within the field of medical natural language processing.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":147646,"compressed_storage_size":18940,"resource_type":"Database","access_policy":"Credentialed","topics":["natural language processing","patient classification"],"source_url":"https://physionet.org/content/phenotype-annotations-mimic/1.20.3/"},{"public_project_uuid":"35f41fde-de0f-4687-9692-f1a7c95e65fe","slug":"ct-ich","version":"1.3.1","core_doi":"10.13026/g1kq-sc84","version_doi":"10.13026/4nae-zg36","is_latest_version":true,"short_description":"Head computed tomography (CT) scans with intracranial hemorrhage (ICH) segmentation, ICH subtypes and skull fracture.","publish_date":"2020-03-10","title":"Computed Tomography Images for Intracranial Hemorrhage Detection and Segmentation","abstract":"After traumatic brain injury (TBI), intracranial hemorrhage (ICH) may occur that could lead to death or disability if it is not accurately diagnosed and treated in a time-sensitive procedure. Currently, Computerized Tomography (CT) scans are examined by radiologists to diagnose intracranial hemorrhage to localize affected regions. In this work, we collected a dataset of 82 CT scans of patients with traumatic brain injury. Intracranial hemorrhage regions in these scans were delineated in each slice by two radiologists. The radiologists also annotated each CT slice for the presence of different types of intracranial hemorrhage and fracture. The CT scans of 75 subjects in NIfTI format are made public in this dataset.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":2950918290,"compressed_storage_size":601120822,"resource_type":"Database","access_policy":"Restricted","topics":["segmentation","computed tomography","nifti","skull fracture","ich","intracranial hemmorhage","ct scans"],"source_url":"https://physionet.org/content/ct-ich/1.3.1/"},{"public_project_uuid":"7274e364-4cfa-4c18-8a92-d3ddcb7b8c4c","slug":"maternal-visceral-adipose","version":"1.0.0","core_doi":"10.13026/9q68-n048","version_doi":"10.13026/p729-7p53","is_latest_version":true,"short_description":"Maternal visceral adipose tissue measurements collected as part of a cohort study of 154 pregnant women.","publish_date":"2020-03-23","title":"Visceral adipose tissue measurements during pregnancy","abstract":"This database is from a cohort study of pregnant women up to 20 weeks of pregnancy and followed until delivery. Measurement of maternal visceral adipose tissue (VAT) was performed during routine obstetric ultrasound. At this time, the biometric evaluation was also performed, and information obtained from prenatal care. Gestational outcomes, including gestational diabetes mellitus (GDM), were obtained through evaluating the patients' medical records at the hospitals where they took place. The data was collected as part of a study that sought to evaluate if maternal VAT could predict GDM at time of delivery. Variables included in this dataset are maternal age, previous DM, blood pressure (in the same day of VAT measurement), VAT (in the periumbilical region), gestational age at time of inclusion, number of pregnancies, level of first fasting glucose, pre-gestational body mass index (BMI). Also, the pregnancy outcomes: gestational age at birth, type of delivery (vaginal or caesarean section), child birth weight and the diagnosis of GDM.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":29499,"compressed_storage_size":11300,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/maternal-visceral-adipose/1.0.0/"},{"public_project_uuid":"2b081428-b719-4a16-bdaf-581f3dbb731f","slug":"mimic3wdb","version":"1.0","core_doi":"10.13026/gs83-bd50","version_doi":"10.13026/c2607m","is_latest_version":true,"short_description":"The MIMIC-III Waveform Database contains numerous physiological signals (including continuous ECG, PPG, ABP, and other signals) and periodic measurements, recorded by bedside patient monitors from about 30,000 patients in intensive care units.","publish_date":"2020-04-07","title":"MIMIC-III Waveform Database","abstract":"The MIMIC-III Waveform Database contains 67,830 record sets for approximately 30,000 ICU patients. Almost all record sets include a waveform record containing digitized signals (typically including ECG, ABP, respiration, and PPG, and frequently other signals) and a “numerics” record containing time series of periodic measurements, each presenting a quasi-continuous recording of vital signs of a single patient throughout an ICU stay (typically a few days, but many are several weeks in duration). A subset of this database contains waveform and numerics records that have been matched and time-aligned with MIMIC-III Clinical Database records.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":7395502909383,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/mimic3wdb/1.0/"},{"public_project_uuid":"03c8718b-688b-4a0c-bfc0-2913d10a33a9","slug":"mimic3wdb-matched","version":"1.0","core_doi":"10.13026/v217-zr73","version_doi":"10.13026/c2294b","is_latest_version":true,"short_description":"Physiological signals (including continuous ECG, PPG, ABP, and other signals) that are associated with patients in the MIMIC-III Clinical Database.","publish_date":"2020-04-07","title":"MIMIC-III Waveform Database Matched Subset","abstract":"The MIMIC-III Waveform Database Matched Subset contains 22,317 waveform records, and 22,247 numerics records, for 10,282 distinct ICU patients. These recordings typically include digitized signals such as ECG, ABP, respiration, and PPG, as well as periodic measurements such as heart rate, oxygen saturation, and systolic, mean, and diastolic blood pressure.
\n\nThis database is a subset of the MIMIC-III Waveform Database, representing those records for which the patient has been identified, and their corresponding clinical records are available in the MIMIC-III Clinical Database.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":2587647437905,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/mimic3wdb-matched/1.0/"},{"public_project_uuid":"f1d3e688-2f90-464c-9270-1e7fed86f40f","slug":"ptb-xl","version":"1.0.0","core_doi":"10.13026/6sec-a640","version_doi":"10.13026/qgmg-0d46","is_latest_version":false,"short_description":"The PTB-XL ECG dataset is a large dataset of 21837 clinical 12-lead ECGs from 18885 patients of 10 second length. The raw signal data has been annotated by up to two cardiologists with 71 different ECG statements and is supplemented by rich metadata.","publish_date":"2020-04-17","title":"PTB-XL, a large publicly available electrocardiography dataset","abstract":"Electrocardiography (ECG) is a key diagnostic tool to assess the cardiac condition of a patient. Automatic ECG interpretation algorithms as diagnosis support systems promise large reliefs for the medical personnel - only on the basis of the number of ECGs that are routinely taken. However, the development of such algorithms requires large training datasets and clear benchmark procedures. In our opinion, both aspects are not covered satisfactorily by existing freely accessible ECG datasets.
\n\nThe PTB-XL ECG dataset is a large dataset of 21837 clinical 12-lead ECGs from 18885 patients of 10 second length. The raw waveform data was annotated by up to two cardiologists, who assigned potentially multiple ECG statements to each record. The in total 71 different ECG statements conform to the SCP-ECG standard and cover diagnostic, form, and rhythm statements. To ensure comparability of machine learning algorithms trained on the dataset, we provide recommended splits into training and test sets. In combination with the extensive annotation, this turns the dataset into a rich resource for the training and the evaluation of automatic ECG interpretation algorithms. The dataset is complemented by extensive metadata on demographics, infarction characteristics, likelihoods for diagnostic ECG statements as well as annotated signal properties.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":5017717909,"compressed_storage_size":3661210833,"resource_type":"Database","access_policy":"Open","topics":["ecg","ptb","ptb-xl","electrocardiography"],"source_url":"https://physionet.org/content/ptb-xl/1.0.0/"},{"public_project_uuid":"ed347333-19aa-4394-9306-ebb004fd4938","slug":"challenge-2020","version":"1.0.1","core_doi":"10.13026/m77n-sx13","version_doi":"10.13026/f4ab-0814","is_latest_version":false,"short_description":"The goal of the 2020 PhysioNet - Computing in Cardiology Challenge is to design and implement a working, open-source algorithm that can automatically identify cardiac abnormalities in 12-lead ECG recordings.","publish_date":"2020-04-23","title":"Classification of 12-lead ECGs: The PhysioNet/Computing in Cardiology Challenge 2020","abstract":"The electrocardiogram (ECG) is a non-invasive representation of the electrical activity of the heart from electrodes placed on the surface of the torso. The standard 12-lead ECG has been widely used to diagnose a variety of cardiac abnormalities such as cardiac arrhythmias, and predicts cardiovascular morbidity and mortality [1]. The early and correct diagnosis of cardiac abnormalities can increase the chances of successful treatments [2]. However, manual interpretation of the electrocardiogram is time-consuming, and requires skilled personnel with a high degree of training [3].
\n\nAutomatic detection and classification of cardiac abnormalities can assist physicians in the diagnosis of the growing number of ECGs recorded. Over the last decade, there have been increasing numbers of attempts to stimulate 12-lead ECG classification. Many of these algorithms seem to have the potential for accurate identification of cardiac abnormalities. However, most of these methods have only been tested or developed in single, small, or relatively homogeneous datasets. The PhysioNet/Computing in Cardiology Challenge 2020 provides an opportunity to address this problem by providing data from a wide set of sources.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":15182,"compressed_storage_size":0,"resource_type":"Challenge","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/challenge-2020/1.0.1/"},{"public_project_uuid":"1bdd27ef-396a-4b4c-b09b-221b45d97a8b","slug":"ptb-xl","version":"1.0.1","core_doi":"10.13026/6sec-a640","version_doi":"10.13026/x4td-x982","is_latest_version":false,"short_description":"The PTB-XL ECG dataset is a large dataset of 21837 clinical 12-lead ECGs from 18885 patients of 10 second length. The raw signal data has been annotated by up to two cardiologists with 71 different ECG statements and is supplemented by rich metadata.","publish_date":"2020-04-24","title":"PTB-XL, a large publicly available electrocardiography dataset","abstract":"Electrocardiography (ECG) is a key diagnostic tool to assess the cardiac condition of a patient. Automatic ECG interpretation algorithms as diagnosis support systems promise large reliefs for the medical personnel - only on the basis of the number of ECGs that are routinely taken. However, the development of such algorithms requires large training datasets and clear benchmark procedures. In our opinion, both aspects are not covered satisfactorily by existing freely accessible ECG datasets.
\n\nThe PTB-XL ECG dataset is a large dataset of 21837 clinical 12-lead ECGs from 18885 patients of 10 second length. The raw waveform data was annotated by up to two cardiologists, who assigned potentially multiple ECG statements to each record. The in total 71 different ECG statements conform to the SCP-ECG standard and cover diagnostic, form, and rhythm statements. To ensure comparability of machine learning algorithms trained on the dataset, we provide recommended splits into training and test sets. In combination with the extensive annotation, this turns the dataset into a rich resource for the training and the evaluation of automatic ECG interpretation algorithms. The dataset is complemented by extensive metadata on demographics, infarction characteristics, likelihoods for diagnostic ECG statements as well as annotated signal properties.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":3187052877,"compressed_storage_size":1842722380,"resource_type":"Database","access_policy":"Open","topics":["ptb-xl","ptb","electrocardiography","ecg"],"source_url":"https://physionet.org/content/ptb-xl/1.0.1/"},{"public_project_uuid":"f31d556a-9830-4c0e-802f-f71613bdd15f","slug":"kinematic-actors-emotions","version":"1.0.0","core_doi":"10.13026/ag2y-yh44","version_doi":"10.13026/ckyh-jf25","is_latest_version":false,"short_description":"1190 kinematic recordings of twenty-two semi-professional actors expressing emotions such as happiness, sadness, anger, fear, and disgust.","publish_date":"2020-04-26","title":"Kinematic dataset of actors expressing emotions","abstract":"We produced a kinematic dataset to assist in recognizing cues from all parts of the body that indicate human emotions (happy, sad, angry, fearful, neutral, and disgust). The present dataset was created using a portable wireless motion capture system. Twenty-two semi-professional actors (50% female) completed performances. A total of 1190 recordings at 125 Hz were collected, consisting of the position and rotation data of 72 anatomical nodes. We hope this dataset will contribute to multiple fields of research and practice, including social neuroscience, psychiatry, computer vision, and biometric and information forensics.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":3736193221,"compressed_storage_size":639767165,"resource_type":"Database","access_policy":"Restricted","topics":["body movement","kinematic data","motion capture","emotion"],"source_url":"https://physionet.org/content/kinematic-actors-emotions/1.0.0/"},{"public_project_uuid":"e3b61396-d865-4728-b690-3678a415d690","slug":"what-is-in-a-note","version":"0.1","core_doi":"10.13026/c2m37c","version_doi":"10.13026/4nen-gv40","is_latest_version":true,"short_description":"Word vectors corresponding to the AMIA 2018 Informatics Summit paper of the same name.","publish_date":"2018-01-08","title":"What's in a Note? Unpacking Predictive Value in Clinical Note Representations","abstract":"Unlocking the full potential of clinical notes within electronic health records presents a challenge for large scale, computational analyses. Many studies that incorporate information from clinical notes do so with simple word-matching approaches and a focused goal of improving performance in a specific task such as mortality prediction. Our project explores representation strategies for clinical text and how to combine many words into larger document-level representations. The code and model shared here are supplements to the AMIA 2018 Informatics Summit paper of the same name.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":144173843,"compressed_storage_size":57861618,"resource_type":"Model","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/what-is-in-a-note/0.1/"},{"public_project_uuid":"9b9d72b3-bf6f-42e4-9e68-429bebc58383","slug":"transformer-synthetic-note","version":"1.0.0","core_doi":"10.13026/h3k0-4982","version_doi":"10.13026/m34x-fq90","is_latest_version":true,"short_description":"Machine learning models that have been trained using MIMIC-III to enable the creation of synthetic discharge summaries.","publish_date":"2020-05-27","title":"Transformer models trained on MIMIC-III to generate synthetic patient notes","abstract":"Natural Language Processing can help to unlock knowledge in the vast troves of unstructured clinical data that are collected during patient care. Patient confidentiality presents a barrier to the sharing and analysis of such data, however, meaning that only small, fragmented and sequestered datasets are available for research. To help side-step this roadblock, we explore the use of Transformer models for the generation of synthetic notes. We demonstrate how models trained on notes from the MIMIC-III clinical database can be used to generate synthetic data with potential to support downstream research studies. We release these trained models to the research community to stimulate further research in this area.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1331005253,"compressed_storage_size":1138234228,"resource_type":"Model","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/transformer-synthetic-note/1.0.0/"},{"public_project_uuid":"4950045a-52d8-4030-b6d3-c400f4766951","slug":"mimic2-iaccd","version":"1.0","core_doi":"10.13026/1r5h-4p36","version_doi":"10.13026/C2NC7F","is_latest_version":true,"short_description":"Dataset extracted from MIMIC-II for a tutorial on effectiveness of indwelling arterial catheters in hemodynamically stable patients with respiratory failure for mortality outcomes.","publish_date":"2016-10-28","title":"Clinical data from the MIMIC-II database for a case study on indwelling arterial catheters","abstract":"This dataset was created for the purpose of a case study in the book: Secondary Analysis of Electronic Health Records, published by Springer in 2016. In particular, the dataset was used throughout Chapter 16 (Data Analysis) by Raffa J. et al. to investigate the effectiveness of indwelling arterial catheters in hemodynamically stable patients with respiratory failure for mortality outcomes. The dataset is derived from MIMIC-II, the publicly-accessible critical care database. It contains summary clinical data and outcomes for 1,776 patients.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":7235657,"compressed_storage_size":6998293,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/mimic2-iaccd/1.0/"},{"public_project_uuid":"b35713bd-f4ac-4482-a3c4-de7530643bff","slug":"hirid","version":"1.0","core_doi":"10.13026/323r-nk04","version_doi":"10.13026/hz5m-md48","is_latest_version":false,"short_description":"The HiRID database contains a large selection of all routinely collected data relating to patient admissions to the Department of Intensive Care Medicine of the Bern University Hospital, Switzerland (ICU).","publish_date":"2020-05-31","title":"HiRID, a high time-resolution ICU dataset","abstract":"HiRID is a freely accessible critical care dataset containing data relating to almost 34 thousand patient admissions to the Department of Intensive Care Medicine of the Bern University Hospital, Switzerland (ICU), an interdisciplinary 60-bed unit admitting >6,500 patients per year. The ICU offers the full range of modern interdisciplinary intensive care medicine for adult patients. The dataset was developed in cooperation between the Swiss Federal Institute of Technology (ETH) Zürich, Switzerland and the ICU.
\n\nThe dataset contains de-identified demographic information and a total of 681 routinely collected physiological variables, diagnostic test results and treatment parameters from almost 34 thousand admissions during the period from January 2008 to June 2016. Data is stored with a uniquely high time resolution of one entry every two minutes.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":18025007962,"compressed_storage_size":17996884979,"resource_type":"Database","access_policy":"Contributor Review","topics":["machine learning","high resolution","icu","intensive care","clinical","critical care"],"source_url":"https://physionet.org/content/hirid/1.0/"},{"public_project_uuid":"ad9541cb-7d92-4bfd-9b59-4dccea242c97","slug":"ludb","version":"1.0.0","core_doi":"10.13026/wdhq-rs83","version_doi":"10.13026/qweb-sr17","is_latest_version":false,"short_description":"ECG signal database that consists of 200 10-second 12-lead records. The boundaries and peaks of P, T waves and QRS complexes were manually annotated by cardiologists. Each record is annotated with the corresponding diagnosis.","publish_date":"2020-06-04","title":"Lobachevsky University Electrocardiography Database","abstract":"Lobachevsky University Electrocardiography Database (LUDB) is an ECG signal database with marked boundaries and peaks of P, T waves and QRS complexes. The database consists of 200 10-second 12-lead ECG signal records representing different morphologies of the ECG signal. The ECGs were collected from healthy volunteers and patients of the Nizhny Novgorod City Hospital No 5 in 2017–2018. The patients had various cardiovascular diseases while some of them had pacemakers. The boundaries of P, T waves and QRS complexes were manually annotated by cardiologists for all 200 records. Also, each record is annotated with the corresponding diagnosis. The database can be used for educational purposes as well as for training and testing algorithms for ECG delineation, i.e. for automatic detection of boundaries and peaks of P, T waves and QRS complexes.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":24760339,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Open","topics":["diagnosis","open database","ecg","delineation","electrocardiography","database"],"source_url":"https://physionet.org/content/ludb/1.0.0/"},{"public_project_uuid":"b03f51d2-bd1f-4a9f-a31a-57465c28c54f","slug":"ecg-spider-clip","version":"1.0.0","core_doi":"10.13026/kc0e-6a32","version_doi":"10.13026/sq6q-zg04","is_latest_version":true,"short_description":"Dataset used for development of an algorithm for on-line anxiety level detection from biosignals.","publish_date":"2020-06-05","title":"Electrocardiogram, skin conductance and respiration from spider-fearful individuals watching spider video clips","abstract":"This project contains electrocardiogram, skin conductance and respiration as raw data (unfiltered, unprocessed) recorded from consented, spider-fearful individuals using the wearable BITalino biosignal measurement device (PLUX – Wireless Biosignals S.A., Lisbon, Portugal) with the sampling rate set to 100 Hz per channel having 10-bit resolution.
\n\nAll subjects were introduced to the exposure procedure and rationale that is mainly based on the principles of the one-session exposure treatment. The main focus of this study was to investigate if the use of an HRV biofeedback intervention could be a promising therapeutic add-on to exposure therapy for specific phobias. Biosignals were collected at the Saarland University (Germany) from July 2017 to July 2018, and there are 57 subjects in total.
\n\nThe dataset may support studies on mental health and psychophysiology related to other anxiety disorders, such as fear of flying, social phobia and posttraumatic stress disorder.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1166666682,"compressed_storage_size":152113922,"resource_type":"Database","access_policy":"Open","topics":["exposure therapy","psychophysiology","biosignals","machine learning","anxiety disorders","wearable sensors","anxiety level detection"],"source_url":"https://physionet.org/content/ecg-spider-clip/1.0.0/"},{"public_project_uuid":"6e8a67dc-4c72-48f6-a139-31b71c99b7d9","slug":"simultaneous-measurements","version":"1.0.0","core_doi":"10.13026/ddzr-dw82","version_doi":"10.13026/chd5-t946","is_latest_version":false,"short_description":"Dataset to support comparison of usability and accuracy from simultaneous measurements collected from 13 subjects including five devices: NeXus-10 MKII, eMotion Faros 360°, Hexoskin Hx1, SOMNOTouch NIBP, Polar RS800 Multi.","publish_date":"2020-06-16","title":"Simultaneous physiological measurements with five devices at different cognitive and physical loads","abstract":"This open database contains physiological measurements from 13 adult and healthy subjects during a standardized experimental setup. The physiology was examined with five different devices (NeXus-10 MKII, eMotion Faros 360°, SOMNOtouch NIBP, Hexoskin Hx1, Polar RS800 Multi), which, in contrast to other studies, were recorded simultaneously. The experiment included a five minute baseline measurement, five minute walking on the treadmill, a five minute cognitive audio test, and a five minute uphill walking on the treadmill. Measurements included electrocardiography (ECG), photoplethysmography, accelerometry, oxygen saturation, respiration, heart rate, heart rate variability, and RR intervals sampled between 1 Hz and 8000 Hz. The dataset was originally generated during an investigation of functionality, accuracy, and usability of several ECG-measurement devices. This resource included the complete raw data files (EDF and HRM) and merged data with a manual expert annotation of heart beats and annotation files with information about the beginning of each experimental phase.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":1791856024,"compressed_storage_size":814748024,"resource_type":"Database","access_policy":"Open","topics":["photoplethysmogram","respiration","multiparameter","temperature","ecg","hrv","movement","holter","noise","accelerometer","heart rate"],"source_url":"https://physionet.org/content/simultaneous-measurements/1.0.0/"},{"public_project_uuid":"f7dd6f30-c90e-4bad-b778-da1ff91982d1","slug":"mmash","version":"1.0.0","core_doi":"10.13026/sntf-5s35","version_doi":"10.13026/cerq-fc86","is_latest_version":true,"short_description":"Multilevel Monitoring of Activity and Sleep in Healthy people (MMASH) dataset provides 24 hours of continuous beat-to-beat heart data, triaxal accelerometer data, sleep quality, physical activity, psychological characteristics and salivary samples.","publish_date":"2020-06-19","title":"Multilevel Monitoring of Activity and Sleep in Healthy People","abstract":"Multilevel Monitoring of Activity and Sleep in Healthy people (MMASH) dataset provides 24 hours of continuous beat-to-beat heart data, triaxial accelerometer data, sleep quality, physical activity and psychological characteristics (i.e., anxiety status, stress events and emotions) for 22 healthy participants. Moreover, saliva bio-markers (i.e.cortisol and melatonin) and activity log were also provided in this dataset. The MMASH dataset will enable researchers to test the correlations between physical activity, sleep quality, and psychological characteristics.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":24675587,"compressed_storage_size":23780300,"resource_type":"Database","access_policy":"Open","topics":["physiological response","melatonin","cortisol","health","circadian rhythm","sleep","psychological response","saliva"],"source_url":"https://physionet.org/content/mmash/1.0.0/"},{"public_project_uuid":"cd1f5291-7e87-433b-a5bf-eec8b04961a1","slug":"kinematic-actors-emotions","version":"2.0.0","core_doi":"10.13026/ag2y-yh44","version_doi":"10.13026/cdb3-8925","is_latest_version":false,"short_description":"1402 kinematic recordings of twenty-two semi-professional actors expressing emotions such as happiness, sadness, anger, fear, disgust, and surprise.","publish_date":"2020-06-23","title":"Kinematic dataset of actors expressing emotions","abstract":"We produced a kinematic dataset to assist in recognizing cues from all parts of the body that indicate human emotions (happy, sad, angry, fearful, , disgust, surprise) and neutral. The present dataset was created using a portable wireless motion capture system. Twenty-two semi-professional actors (50% female) completed performances. A total of 1402 recordings at 125 Hz were collected, consisting of the position and rotation data of 72 anatomical nodes. We hope this dataset will contribute to multiple fields of research and practice, including social neuroscience, psychiatry, computer vision, and biometric and information forensics.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":4356276808,"compressed_storage_size":745413393,"resource_type":"Database","access_policy":"Restricted","topics":["motion capture","emotion","body movement","kinematic data"],"source_url":"https://physionet.org/content/kinematic-actors-emotions/2.0.0/"},{"public_project_uuid":"cc675606-124a-4538-a65c-5ec7e1cff358","slug":"kinematic-actors-emotions","version":"2.1.0","core_doi":"10.13026/ag2y-yh44","version_doi":"10.13026/kg8b-1t49","is_latest_version":true,"short_description":"1402 kinematic recordings of twenty-two semi-professional actors expressing emotions such as happiness, sadness, anger, fear, disgust, and surprise.","publish_date":"2020-07-07","title":"Kinematic dataset of actors expressing emotions","abstract":"We produced a kinematic dataset to assist in recognizing cues from all parts of the body that indicate human emotions (happy, sad, angry, fearful, , disgust, surprise) and neutral. The present dataset was created using a portable wireless motion capture system. Twenty-two semi-professional actors (50% female) completed performances. A total of 1402 recordings at 125 Hz were collected, consisting of the position and rotation data of 72 anatomical nodes. We hope this dataset will contribute to multiple fields of research and practice, including social neuroscience, psychiatry, computer vision, and biometric and information forensics.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":4356277833,"compressed_storage_size":745414079,"resource_type":"Database","access_policy":"Restricted","topics":["body movement","emotion","motion capture","kinematic data"],"source_url":"https://physionet.org/content/kinematic-actors-emotions/2.1.0/"},{"public_project_uuid":"3da9967b-7bd1-4a62-9439-17c7f720eb91","slug":"butqdb","version":"1.0.0","core_doi":"10.13026/qcnm-a155","version_doi":"10.13026/kah4-0w24","is_latest_version":true,"short_description":"The database is intended for the development and objective comparison of algorithms designed to assess the quality of ECG records. It also enables objective comparison of results between authors.","publish_date":"2020-07-23","title":"Brno University of Technology ECG Quality Database (BUT QDB)","abstract":"BUT QDB is a database created by the cardiology team at the Department of Biomedical Engineering, Brno University of Technology, for the purpose of evaluating ECG quality. The data comprises 18 long-term recordings of single-lead ECGs and associated 3-axis accelerometer data, collected from 15 subjects (9 female, 6 male) aged between 21 to 83 years. Recordings were carried out between August 2018 and October 2019 while the subjects were undertaking ordinary everyday activities ("free-living conditions"). Data was collected using a mobile ECG and accelerometer (Bittium Faros 180) with a sampling frequency of 1,000 Hz for ECG signals and 100 Hz for accelerometer signals. The minimum length of recording is 24 hours.
\n\nThree signals were fully annotated in terms of ECG signal quality. The remaining 15 signals were annotated in two selected segments, each of 20 minutes in duration. Furthermore, five additional segments of poor signal quality were also annotated. Signal quality was classified as follows: Class 1 indicates that all ECG significant waveforms (P wave, T wave and QRS complex) are clearly visible and their onsets and offsets can be detected reliably; Class 2 indicates that the noise level is increased and ECG significant points cannot be reliably detected, but the signal enables reliable QRS detection; Class 3 indicates that QRS complexes cannot be detected reliably and the signal is unsuitable for any further analysis.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":4481644508,"compressed_storage_size":2394096389,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/butqdb/1.0.0/"},{"public_project_uuid":"e2dc43ab-f0be-4dd3-a12a-18a1088984fa","slug":"bhx-brain-bounding-box","version":"1.1","core_doi":"10.13026/7br9-xy10","version_doi":"10.13026/9cft-hg92","is_latest_version":true,"short_description":"The first version of this dataset was made available in the forum of Kaggle competition 'RSNA Intracranial Hemorrhage Detection' (v1.0). Then minor corrections were implemented (v1.1).","publish_date":"2020-07-29","title":"Brain Hemorrhage Extended (BHX): Bounding box extrapolation from thick to thin slice CT images","abstract":"BHX is a public available dataset with bounding box annotations for 5 types of acute hemorrhage as an extension of the qure.ai CQ500 dataset. This dataset intends to provide data resources to help advance hemorrhage detection towards machine learning localization tasks.
\n\nKey Points:
\n\nBHX contains up to 39,668 bounding boxes in 23,409 images annotated for hemorrhage, out of a total of ~170k images from qure.ai CQ500 dataset.
\n\tThe dataset was built through an efficient method to obtain automatic annotated images (thin slices) from sparse initial labeling (thick slices).
\n\tWe hope this dataset can contribute to further develop and validate machine learning algorithms, ultimately helping to diagnose intracranial hemorrhage.
\n\tThis dataset of electrodermal activity was collected from 11 healthy volunteer subjects who were awake and at rest in seated position. The activity was recorded from each subject's non-dominant hand for one hour at 256 Hz. Electrodermal activity measures changing electrical conductance of the skin as an indicator of sweat gland activity [1]. Sweat glands are a primitive part of the fight-or-flight response [1]. These data were collected as part of a larger study to understand and build computational models for autonomic nervous system activity (including electrodermal activity) with approval from the Massachusetts Institute of Technology Committee on the Use of Humans as Experimental Subjects (COUHES).
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":71074181,"compressed_storage_size":1237134,"resource_type":"Database","access_policy":"Credentialed","topics":["sweat gland activity","sympathetic nervous system","autonomic nervous system","electrodermal activity"],"source_url":"https://physionet.org/content/electrodermal-activity/1.0/"},{"public_project_uuid":"0759de72-40a6-4e1a-ad4e-72d919a99322","slug":"siena-scalp-eeg","version":"1.0.0","core_doi":"10.13026/s309-a395","version_doi":"10.13026/5d4a-j060","is_latest_version":true,"short_description":"The database consists of EEG recordings of 14 epileptic patients acquired at the Unit of Neurology and Neurophysiology of the University of Siena. Subjects include 9 males (ages 25-71) and 5 females (ages 20-58).","publish_date":"2020-08-11","title":"Siena Scalp EEG Database","abstract":"The database consists of EEG recordings of 14 patients acquired at the Unit of Neurology and Neurophysiology of the University of Siena. Subjects include 9 males (ages 25-71) and 5 females (ages 20-58). Subjects were monitored with a Video-EEG with a sampling rate of 512 Hz, with electrodes arranged on the basis of the international 10-20 System. Most of the recordings also contain 1 or 2 EKG signals. The diagnosis of epilepsy and the classification of seizures according to the criteria of the International League Against Epilepsy were performed by an expert clinician after a careful review of the clinical and electrophysiological data of each patient.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":21799417249,"compressed_storage_size":14001243214,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/siena-scalp-eeg/1.0.0/"},{"public_project_uuid":"c77c053f-06d8-4e34-a688-7737c8a70733","slug":"mimiciv","version":"0.3","core_doi":"10.13026/07hj-2a80","version_doi":"10.13026/a2mm-bn44","is_latest_version":false,"short_description":"Database of patients admitted to the Beth Israel Deaconess Medical Center","publish_date":"2020-08-14","title":"MIMIC-IV","abstract":"Retrospectively collected medical data has the opportunity to improve patient care through knowledge discovery and algorithm development. Broad reuse of medical data is desirable for the greatest public good, but data sharing must be done in a manner which protects patient privacy. The Medical Information Mart for Intensive Care (MIMIC)-III database provided critical care data for over 40,000 patients admitted to intensive care units at the Beth Israel Deaconess Medical Center (BIDMC). Importantly, MIMIC-III was deidentified, and patient identifiers were removed according to the Health Insurance Portability and Accountability Act (HIPAA) Safe Harbor provision. MIMIC-III has been integral in driving large amounts of research in clinical informatics, epidemiology, and machine learning. Here we present MIMIC-IV, an update to MIMIC-III, which incorporates contemporary data and improves on numerous aspects of MIMIC-III. MIMIC-IV adopts a modular approach to data organization, highlighting data provenance and facilitating both individual and combined use of disparate data sources. MIMIC-IV is intended to carry on the success of MIMIC-III and support a broad set of applications within healthcare.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":7340949009,"compressed_storage_size":7275244007,"resource_type":"Database","access_policy":"Credentialed","topics":["intensive care unit","machine learning","critical care","mimic"],"source_url":"https://physionet.org/content/mimiciv/0.3/"},{"public_project_uuid":"bb264cd3-dae9-40ce-aa9f-06370bc547af","slug":"mimiciv","version":"0.4","core_doi":"10.13026/07hj-2a80","version_doi":"10.13026/a3wn-hq05","is_latest_version":false,"short_description":"Database of patients admitted to the Beth Israel Deaconess Medical Center","publish_date":"2020-08-14","title":"MIMIC-IV","abstract":"Retrospectively collected medical data has the opportunity to improve patient care through knowledge discovery and algorithm development. Broad reuse of medical data is desirable for the greatest public good, but data sharing must be done in a manner which protects patient privacy. The Medical Information Mart for Intensive Care (MIMIC)-III database provided critical care data for over 40,000 patients admitted to intensive care units at the Beth Israel Deaconess Medical Center (BIDMC). Importantly, MIMIC-III was deidentified, and patient identifiers were removed according to the Health Insurance Portability and Accountability Act (HIPAA) Safe Harbor provision. MIMIC-III has been integral in driving large amounts of research in clinical informatics, epidemiology, and machine learning. Here we present MIMIC-IV, an update to MIMIC-III, which incorporates contemporary data and improves on numerous aspects of MIMIC-III. MIMIC-IV adopts a modular approach to data organization, highlighting data provenance and facilitating both individual and combined use of disparate data sources. MIMIC-IV is intended to carry on the success of MIMIC-III and support a broad set of applications within healthcare.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":7242834647,"compressed_storage_size":7177596150,"resource_type":"Database","access_policy":"Credentialed","topics":["machine learning","critical care","intensive care unit","mimic"],"source_url":"https://physionet.org/content/mimiciv/0.4/"},{"public_project_uuid":"05160c7f-0826-4d98-bdeb-6028c5e3c8f2","slug":"r-deco","version":"1.0.0","core_doi":"10.13026/gxxm-xw77","version_doi":"10.13026/x6j7-sp58","is_latest_version":true,"short_description":"An open-source Matlab based graphical user interface for the detection and correction of R-peaks.","publish_date":"2020-09-09","title":"R-DECO: An open-source Matlab based graphical user interface for the detection and correction of R-peaks","abstract":"Many of the existing ECG toolboxes focus on the derivation of heart rate variability features from RR-intervals. By doing so, they assume correct detection of the QRS-complexes. However, it is highly likely that not all detections are correct. Therefore, it is recommended to visualize the actual R-peak positions in the ECG signal and allow manual adaptations.
\n\nAs a solution, we present R-DECO, an easy-to-use graphical user interface for the detection and correction of R-peaks. Within R-DECO, the R-peaks are detected by an adaptation of the Pan-Tompkins algorithm. Instead of using all the pre-processing steps of the latter algorithm, the proposed algorithm uses only an envelope-based procedure to flatten the ECG and enhance the QRS-complexes. The algorithm obtained an overall sensitivity of 99.60% and positive predictive value of 99.69% on the MIT/BIH arrhythmia database. Additionally, R-DECO includes support for several input data formats for ECG signals, three basic filters, the possibility to load other R-peak locations and intuitive methods to correct ectopic, wrong, or missed heartbeats. All functionalities can be accessed via the graphical user interface and the analysis results can be exported as Matlab or Excel files.
\n\nThrough its easy-to-use-graphical user interface, R-DECO allows both clinicians and researchers to use all functionalities, without previous knowledge. R-DECO has been developed on Windows 10 using Matlab 2018a. GUIDE, MATLAB’s GUI development environment, was used to design the GUI of R-DECO. It has been tested and proven to be functional for older Matlab versions going back to Matlab 2016a.
","license":{"name":"GNU General Public License version 3"},"dua":null,"main_storage_size":952916,"compressed_storage_size":676210,"resource_type":"Software","access_policy":"Open","topics":["graphical user interface","algorithms and analysis of algorithms","signal processiong","visual analysis"],"source_url":"https://physionet.org/content/r-deco/1.0.0/"},{"public_project_uuid":"59c5819b-ff24-43a6-9070-d1710ba05ae3","slug":"heart-failure-zigong","version":"1.0","core_doi":"10.13026/wd7p-qy77","version_doi":"10.13026/3pr8-jx39","is_latest_version":false,"short_description":"Dataset comprising hospital-level data on patients who were admitted with heart failure to Zigong Fourth People’s Hospital, Sichuan, China between 2016 and 2019","publish_date":"2020-09-09","title":"Hospitalized patients with heart failure: integrating electronic healthcare records and external outcome data","abstract":"Heart failure is a common reason for hospitalization in the elderly and it is associated with significant mortality and morbidity. To facilitate epidemiological studies of heart failure, there is a need for high quality datasets to be made available to researchers. While several heart failure datasets have been established in Western countries, there is a paucity of data available from China. Understanding differences in patient populations and healthcare systems between China and other countries is important in providing optimal care. To help address this issue, we created a retrospective heart failure dataset using electronic health data collected from patients who were admitted to a hospital in Sichuan, China between 2016 and 2019. The dataset includes 168 variables for 2,008 patients with heart failure.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":1499242,"compressed_storage_size":398787,"resource_type":"Database","access_policy":"Restricted","topics":["heart failure","electronic health record","china"],"source_url":"https://physionet.org/content/heart-failure-zigong/1.0/"},{"public_project_uuid":"ac2d6f09-2051-4272-83f0-163109577d49","slug":"egd-cxr","version":"1.0.0","core_doi":"10.13026/nrbg-cs59","version_doi":"10.13026/qfdz-zr67","is_latest_version":true,"short_description":"This dataset was a collected using an eye tracking system while a radiologist interpreted and read 1,083 public CXR images. The dataset contains the following aligned modalities: image, transcribed report text, dictation audio and eye gaze data.","publish_date":"2020-09-13","title":"Eye Gaze Data for Chest X-rays","abstract":"We created a rich multimodal dataset for the Chest X-Ray (CXR) domain. The data was collected using an eye tracking system while a radiologist interpreted and read 1,083 public CXR images. The dataset contains the following aligned modalities: image, transcribed report text, dictation audio and eye gaze data. We hope this dataset can contribute to various fields of research with applications in machine learning such as deep learning explainability, multi-modal fusion, disease classification, and automated radiology report generation to name a few. The images were selected from the MIMIC-CXR Database and were associated with studies from 1,038 subjects (female: 495, male: 543) who had age range 20 - 80 years old.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":4396780185,"compressed_storage_size":1087683124,"resource_type":"Database","access_policy":"Credentialed","topics":["chest","cxr","audio","convolutional network","eye tracking","radiology","multimodal","deep learning","explainability","machine learning","heatmap","chest x-ray"],"source_url":"https://physionet.org/content/egd-cxr/1.0.0/"},{"public_project_uuid":"386181f2-90d7-4a35-bb63-2c78f662c807","slug":"cxr-phone","version":"1.0.0","core_doi":"10.13026/2grc-ny75","version_doi":"10.13026/7b2j-nq93","is_latest_version":true,"short_description":"Smartphone-captured CXR images including photographs taken from MIMIC-CXR and CheXpert, photographs taken by resident doctors, and photographs taken with different devices.","publish_date":"2020-09-28","title":"Smartphone-Captured Chest X-Ray Photographs","abstract":"Recent research has applied deep learning imaging techniques to detect pulmonary pathology in chest X-rays (CXR) and achieved performance comparable with radiologists on specific lung abnormalities. Industry has managed to deploy such artificial intelligence (AI) solutions in on mobile devices to help clinicians improve their workflow. However, the feasibility of applying these modern AI algorithms to CXR photographs has yet to be evaluated. In this project, 6,453 smartphone photographs were taken from the frontal-view CXR images from two publicly available databases, MIMIC-CXR and CheXpert. We constructed four derivative CXR photograph datasets (Photo-MMC, Photo-CXP, Photo-MED, and Photo-DEV) including photographs taken by several resident doctors and photographs taken with different devices.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":797694641,"compressed_storage_size":795645013,"resource_type":"Database","access_policy":"Restricted","topics":["smartphone","photograph","cxr"],"source_url":"https://physionet.org/content/cxr-phone/1.0.0/"},{"public_project_uuid":"76d79284-9c1e-48a2-a54b-ce4754bb3c3f","slug":"mimic-seqex","version":"1.0.0","core_doi":"10.13026/6jx9-cy34","version_doi":"10.13026/n2v5-5b32","is_latest_version":true,"short_description":"MIMIC-III data converted into TensorFlow SequenceExample format, for use in modeling pipelines.","publish_date":"2020-09-29","title":"MIMIC-III - SequenceExamples for TensorFlow modeling","abstract":"This dataset contains TensorFlow SequenceExamples derived from patient records in MIMIC-III, a freely available set of deidentified medical records from critical care patients at Beth Israel Deaconess Medical Center. Each SequenceExample converts data from an individual patient encounter and any previous encounters into a set of timestamped “feature lists” describing the patient history up to a certain time, beyond which predictions can be made. These data are suitable for direct input into TensorFlow modeling pipelines, and include labels for inpatient mortality and discharge diagnosis codes for each encounter. The intent of this release is to provide a preprocessed, ready-to-use version of MIMIC-III to support and enable reproducible machine learning research for electronic health records.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":35725904523,"compressed_storage_size":34460593223,"resource_type":"Database","access_policy":"Credentialed","topics":["deep learning","tensorflow","machine learning","sequence modeling"],"source_url":"https://physionet.org/content/mimic-seqex/1.0.0/"},{"public_project_uuid":"1a5abb55-063a-4643-a032-740073dabdd8","slug":"music-motion-2012","version":"1.0","core_doi":"10.13026/j69z-2802","version_doi":"10.13026/4h9d-gf10","is_latest_version":true,"short_description":"How and why does music make us move? This dataset was collected as part of a project to investigate how music influences small magnitude motion observed when people try to stand still.","publish_date":"2020-10-05","title":"MICRO Motion capture data from groups of participants standing still to auditory stimuli (2012)","abstract":"This dataset comprises head motion observations collected as part of an experiment in which a group of people were asked to stand still for 6 minutes, with the first 3 minutes in silence, followed by 3 minutes with music. Head motion was captured in units of mm at 100Hz using a Qualisys infra-red optical system. The experiment was carried out at the University of Oslo on March 8th 2012 from a total of 113 participants. Code to read and process these files is available. The paper corresponding to the work is Jensenius et al., "The Musical Influence on People's Micromotion when Standing Still in Groups", Proceedings of the 14th Sound and Music Computing Conference (2017).
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":119868882,"compressed_storage_size":58482611,"resource_type":"Database","access_policy":"Open","topics":["movement","music","sound"],"source_url":"https://physionet.org/content/music-motion-2012/1.0/"},{"public_project_uuid":"da497509-5dd9-4632-a406-60ec56b047ac","slug":"heart-failure-zigong","version":"1.1","core_doi":"10.13026/wd7p-qy77","version_doi":"10.13026/q712-yv80","is_latest_version":false,"short_description":"Dataset comprising hospital-level data on patients who were admitted with heart failure to Zigong Fourth People’s Hospital, Sichuan, China between 2016 and 2019","publish_date":"2020-10-09","title":"Hospitalized patients with heart failure: integrating electronic healthcare records and external outcome data","abstract":"Heart failure is a common reason for hospitalization in the elderly and it is associated with significant mortality and morbidity. To facilitate epidemiological studies of heart failure, there is a need for high quality datasets to be made available to researchers. While several heart failure datasets have been established in Western countries, there is a paucity of data available from China. Understanding differences in patient populations and healthcare systems between China and other countries is important in providing optimal care. To help address this issue, we created a retrospective heart failure dataset using electronic health data collected from patients who were admitted to a hospital in Sichuan, China between 2016 and 2019. The dataset includes 168 variables for 2,008 patients with heart failure.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":4201890,"compressed_storage_size":637452,"resource_type":"Database","access_policy":"Restricted","topics":["china","electronic health record","heart failure"],"source_url":"https://physionet.org/content/heart-failure-zigong/1.1/"},{"public_project_uuid":"52b3d6d0-e310-4ebc-b66f-c8faf3c86d0e","slug":"cded","version":"1.0.0","core_doi":"10.13026/bpnv-4b88","version_doi":"10.13026/t92y-x219","is_latest_version":false,"short_description":"Type 2 diabetes increases risk for cerebrovascular disease, cognitive and mobility decline in older people. This project evaluated relationship between diabetes, inflammation cerebrovascular reactivity and functional outcomes.","publish_date":"2020-10-09","title":"Cerebromicrovascular Disease in Elderly with Diabetes","abstract":"This dataset was collected as part of a prospective observational study to evaluate the effects of type 2 diabetes mellitus (DM) on cerebral vasoregulation, perfusion and functional outcomes, measured by blood flow responses to hypocapnia and hypercapnia, Valsalva maneuver, head-up tilt, and sit-to-stand test. The dataset comprises of observations from 69 diabetic and control participants (aged 55 to 75 years) with continuous measurements of cerebral blood flow using transcranial Doppler and MRI (magnetic resonance imaging), heart rate, blood pressure, and respiratory parameters, balance, walking, laboratory and retinopathy measures at baseline, and 41 subjects who completed the two-years of follow-up. Regional gray, white matter and cerebrospinal fluid volumes were quantified using a segmentation method applied on T1- and T2- weighted images and perfusion maps, using a continuous arterial spin labeling (CASL) images at 3 Tesla MRI. White matter integrity was determined from fluid attenuated inversion recovery (FLAIR) and diffusion tensor imaging (DTI) MRI. Dynamics of cerebral vasoregulation to CO2 challenge and orthostatic stress were measured using Transcranial Doppler ultrasound (TCD).
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":3941742930,"compressed_storage_size":2211624159,"resource_type":"Database","access_policy":"Open","topics":["inflammation","cerebromiscovascular disease","gait and balance","type 2 diabetes","perfusion mri","brain perfusion","cerebral vasoreactivity","brain atrophy"],"source_url":"https://physionet.org/content/cded/1.0.0/"},{"public_project_uuid":"753ee4b6-29b6-4d7b-b03b-fc9c795dca56","slug":"emer-complaint-gout","version":"1.0","core_doi":"10.13026/8zq3-b355","version_doi":"10.13026/96v3-dw72","is_latest_version":true,"short_description":"A corpus of chief complaints tagged with predicted gout flare status and chart reviewed gout flare status. Ideal for input to masked language model training to supplement lengthy clinical text notes.","publish_date":"2020-10-19","title":"Gout Emergency Department Chief Complaint Corpora","abstract":"The Gout Emergency Department Chief Complaint Corpora (GED3C) consists of 2 corpora of free text triage nurse chief complaints (up to 282 characters in length) collected from 2019 to 2020 at an academic medical center in the Deep South. The smaller corpus "GOUT-CC-2019-CORPUS" consists of 300 chief complaints from 2019 selected by the presence of the keyword "gout". The larger corpus "GOUT-CC-2020-CORPUS" contains 8037 chief complaints collected from a single month in 2020. No other selection criteria (gout or otherwise) was used to generate GOUT-CC-2020-CORPUS, making this corpus representative for medical conditions of interest in the underlying urban Black majority Emergency Department patient population.
\n\nWe anticipate this corpus being useful in the development of Emergency Department alerting algorithms for gout. Both GOUT-CC-2019-CORPUS and GOUT-CC-2020-CORPUS were annotated with respect to predicted gout flare status as determined retrospectively by manual review of the chief complaint. A subset of patients with these chief complaints underwent chart review by rheumatologists to verify gout flare status guided by the Gaffo criteria. These corpora may be useful to researchers at other institutions who want to develop or validate existing alerting algorithms at a 2nd institution. These corpora are also available for use in masked language model development. Its terse, abbreviation-rich content sets it apart from more lengthy clinical text and to our knowledge this is the only chief complaint corpus publicly available.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":3916291,"compressed_storage_size":1356221,"resource_type":"Database","access_policy":"Restricted","topics":["emergency department","nlp","gout"],"source_url":"https://physionet.org/content/emer-complaint-gout/1.0/"},{"public_project_uuid":"b4ce2530-6a5c-4b81-a4cf-6656c3a91593","slug":"antimicrobial-resistance-uti","version":"1.0.0","core_doi":"10.13026/hdza-eh80","version_doi":"10.13026/se6w-f455","is_latest_version":true,"short_description":"AMR-UTI is a freely accessible dataset, derived from electronic health record (EHR) information on over 100,000 urinary tract infections (UTI) treated at Massachusetts General Hospital and Brigham & Women's Hospital in Boston, MA, USA.","publish_date":"2020-11-04","title":"AMR-UTI: Antimicrobial Resistance in Urinary Tract Infections","abstract":"AMR-UTI is a freely accessible dataset, derived from electronic health record (EHR) information on over 80,000 patients with urinary tract infections (UTI) treated at Massachusetts General Hospital and Brigham & Women's Hospital in Boston, MA, USA between 2007 and 2016.
\n\nEach observation in the dataset corresponds to a urine specimen sent to the clinical microbiology laboratory to assess for antimicrobial resistance (AMR). Each observation includes (1) the antimicrobial susceptibility profile, (2) the empiric antibiotic treatment decision, which is made without knowledge of the susceptibility testing results, and (3) patient and specimen features useful for prediction of AMR.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":378589215,"compressed_storage_size":16377672,"resource_type":"Database","access_policy":"Credentialed","topics":["machine learning","antibiotic resistance","clinical decision support","causal inference","policy learning","antimicrobial resistance","urinary tract infection"],"source_url":"https://physionet.org/content/antimicrobial-resistance-uti/1.0.0/"},{"public_project_uuid":"b43c30ca-c76a-43a3-88c2-3ba3dee444b7","slug":"picdb","version":"1.1.0","core_doi":"10.13026/c9gh-n952","version_doi":"10.13026/32x9-wv38","is_latest_version":true,"short_description":"PIC (Paediatric Intensive Care) is a large paediatric-specific, single-centre, bilingual database comprising information relating to children admitted to critical care units at a large children’s hospital in China.","publish_date":"2020-11-12","title":"Paediatric Intensive Care database","abstract":"The Paediatric Intensive Care (PIC) database is a large paediatric-specific, single-center, bilingual database comprising information relating to children admitted to critical care units at a large children’s hospital in China. The database is de-identified and includes vital sign measurements, medications, laboratory measurements, fluid balance recordings, diagnostic codes, demographic information, and more. The data are publicly available after credentialing, which includes completion of a training course on research with human subjects and signing of a data use agreement mandating responsible handling of the data and adherence to the principle of collaborative research. The PIC database builds upon the success of the widely used MIMIC (Medical Information Mart for Intensive Care) database, and extends the approach into the field of paediatric critical care. The database has many unique characteristics which support academic and industrial research including the development of machine learning algorithms, clinical decision support tools, quality improvement initiatives, and international data sharing.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1522722939,"compressed_storage_size":400745498,"resource_type":"Database","access_policy":"Credentialed","topics":["pediatrics","critical care","intensive care","natural language processing"],"source_url":"https://physionet.org/content/picdb/1.1.0/"},{"public_project_uuid":"cd997daa-b9b4-4ae3-9301-9f6ebe8e92df","slug":"ninfea","version":"1.0.0","core_doi":"10.13026/393a-ky74","version_doi":"10.13026/c4n5-3b04","is_latest_version":true,"short_description":"Open dataset featuring non-invasive electrophysiological recordings, fetal pulsed-wave Doppler and maternal respiration signals. It provides a ground truth on the fetal heart activity when an invasive scalp lead is unavailable.","publish_date":"2020-11-12","title":"NInFEA: Non-Invasive Multimodal Foetal ECG-Doppler Dataset for Antenatal Cardiology Research","abstract":"The development of algorithms for the extraction of the foetal ECG (fECG) from non-invasive recordings is hampered by the lack of publicly-available reference datasets, which could be used to benchmark different algorithms while providing a ground truth on the foetal heart activity when an invasive scalp lead is unavailable. By enriching the electrophysiological recordings with simultaneous multimodal signals, these datasets could also help the investigation of the foetal cardiac physiology, providing ground truth for the analysis in early pregnancy, when the fECG is not directly accessible.
\n\nThe Non-Invasive Multimodal Foetal ECG-Doppler Dataset for Antenatal Cardiology Research (NInFEA) is the first open-access dataset featuring simultaneous non-invasive electrophysiological recordings, fetal pulsed-wave Doppler (PWD) and maternal respiration signals. The dataset includes 60 entries from 39 voluntary pregnant women, between the 21st and the 27th week of gestation. Every entry is composed of 27 electrophysiological channels (2048 Hz, 22 bits, acquired by means of the TMSi Porti7 system), maternal respiration signal (through a resistive thoracic belt), synchronised foetal trans-abdominal PWD and clinical annotations provided by expert clinicians at the time of the signal collection.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":2094600110,"compressed_storage_size":831389363,"resource_type":"Database","access_policy":"Open","topics":["foetus","pwd","doppler","foetal ecg","maternal ecg","pwd envelope","non-invasive","cardiology","early pregnancy","ecg","antenatal","fecg"],"source_url":"https://physionet.org/content/ninfea/1.0.0/"},{"public_project_uuid":"8f6c5360-8f1a-4a79-8ee0-34687fd9a721","slug":"heart-failure-zigong","version":"1.2","core_doi":"10.13026/wd7p-qy77","version_doi":"10.13026/8a9e-w734","is_latest_version":false,"short_description":"Dataset comprising hospital-level data on patients who were admitted with heart failure to Zigong Fourth People’s Hospital, Sichuan, China between 2016 and 2019","publish_date":"2020-12-02","title":"Hospitalized patients with heart failure: integrating electronic healthcare records and external outcome data","abstract":"Heart failure is a common reason for hospitalization in the elderly and it is associated with significant mortality and morbidity. To facilitate epidemiological studies of heart failure, there is a need for high quality datasets to be made available to researchers. While several heart failure datasets have been established in Western countries, there is a paucity of data available from China. Understanding differences in patient populations and healthcare systems between China and other countries is important in providing optimal care. To help address this issue, we created a retrospective heart failure dataset using electronic health data collected from patients who were admitted to a hospital in Sichuan, China between 2016 and 2019. The dataset includes 168 variables for 2,008 patients with heart failure.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":2028060,"compressed_storage_size":480423,"resource_type":"Database","access_policy":"Restricted","topics":["electronic health record","heart failure","china"],"source_url":"https://physionet.org/content/heart-failure-zigong/1.2/"},{"public_project_uuid":"04a468b6-39d3-46c2-959e-fe28eda4027c","slug":"hospital-activity-bosch","version":"1.0","core_doi":"10.13026/g6qh-pf06","version_doi":"10.13026/yrz5-5909","is_latest_version":true,"short_description":"Measurements of physical activity with wrist-worn Bosch sensor platform to test predictive performance for the duration of hospitalization and readmission in 58 patients with acute illnesses in internal medicine","publish_date":"2020-12-03","title":"In-hospital physical activity measured with a new Bosch accelerometer sensor system","abstract":"Extended hospital stays and readmissions are associated with high costs and burden of disease. Effective ways to predict patient outcomes, select those at risk of adverse outcomes, and to prevent these accordingly, are lacking. Physical activity measured by accelerometry is a promising surrogate parameter of health, but it is mainly evaluated in highly selected patient groups and with few algorithms. We used a new wrist-worn Bosch sensor platform to measure various characteristics of physical activity in 58 patients of various ages treated on internal medicine wards over a period of 30 days, with a goal of developing algorithms to predict duration of hospitalization and 30-day readmission. New algorithms are required to increase the accuracy of the predictions and to understand the significance of physical activity in hospitalized patients. To facilitate research in this area, and to allow the public to explore these issues, all data generated in the study are published here as a public dataset.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":63482660973,"compressed_storage_size":12045884631,"resource_type":"Database","access_policy":"Restricted","topics":["prediction","acute illness","accelerometry","accelerometer","hospitalization","readmission"],"source_url":"https://physionet.org/content/hospital-activity-bosch/1.0/"},{"public_project_uuid":"9b29571d-b901-488a-bf0d-c7872ea6062f","slug":"maternal-ultrasound-nutrition","version":"1.0.0","core_doi":"10.13026/zedg-p783","version_doi":"10.13026/hfks-3d71","is_latest_version":true,"short_description":"Dataset collected as part of a prospective study in which abdominal maternal fat tissue measurements were compared with outcomes during hospitalization for labor and delivery.","publish_date":"2020-12-04","title":"Maternal fat ultrasound measurement and nutritional assessment during pregnancy: A dataset centered in gestational outcomes","abstract":"This dataset was collected as part of a prospective study in which abdominal maternal fat tissues ultrasound measurements were compared with outcomes during hospitalization for labor and delivery. Data was collected in a low-risk outpatient clinic from October 2016 to December 2017 in Porto Alegre, Brazil. Of the 272 patients initially included, 61 (22%) were lost to follow-up, resulting in a final sample of 211 cases.
At baseline, a routine obstetric ultrasound was performed with a maternal epigastric fat measurement conducted in all included cases. Two assessments were performed: the epigastric maternal visceral adipose tissue (preperitoneal m-VAT) and the epigastric maternal subcutaneous adipose tissue (preperitoneal m-SAT). Cases included before 20 weeks had an additional periumbilical fat assessment performed; this included: the periumbilical maternal visceral adipose tissue (periumbilical m-VAT) and the periumbilical maternal subcutaneous adipose tissue (periumbilical m-SAT). Data regarding anthropometric, nutritional and demographic characteristics were assessed at baseline. Pregnancy outcomes were determined through hospital charts in order to establish pregnancy, labor and newborn conditions.
\n\nThe dataset has 116 variables that include demographic, ultrasound, nutritional and laboratory characteristics measured at baseline. Likewise, it presents data related to the birth process and newborn conditions during the postpartum period. The pregnancy abnormalities data obtained retrospectively are highlighted. The dataset may be helpful for future studies that seek to assess the predictive capacity of maternal abdominal fat for abnormal pregnancy outcomes.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":167068,"compressed_storage_size":33271,"resource_type":"Database","access_policy":"Credentialed","topics":["pregnancy","abdominal","ultrasound"],"source_url":"https://physionet.org/content/maternal-ultrasound-nutrition/1.0.0/"},{"public_project_uuid":"24658af7-cc7d-42f0-bbe2-6193044cee4c","slug":"music-motion-2015","version":"1.0","core_doi":"10.13026/gdxk-7395","version_doi":"10.13026/dfv0-sb95","is_latest_version":true,"short_description":"How and why does music make us move? This project investigates how music influences small magnitude motion observed when people try to stand still.","publish_date":"2020-12-08","title":"MICRO Motion capture data from groups of participants standing still to auditory stimuli (2015)","abstract":"This project contains head movement data recorded from groups of participants asked to stand as still as possible and presented with a series of auditory stimuli. Data was collected in units of mm with a Qualisys motion capture system at 100Hz. Data was collected at the University of Oslo on March 12th 2015 from a total of 108 participants. Code to read and process these files has been made publicly available.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":183703491,"compressed_storage_size":67309527,"resource_type":"Database","access_policy":"Open","topics":["movement","sound","music cognition"],"source_url":"https://physionet.org/content/music-motion-2015/1.0/"},{"public_project_uuid":"44d23c58-59f7-47e7-a881-fa6c581bbeea","slug":"semeval2015","version":"1.0","core_doi":"10.13026/96re-n451","version_doi":"10.13026/xb00-c765","is_latest_version":false,"short_description":"This is the dataset for SemEval 2015, task 4, Analysis of Clinical Text","publish_date":"2014-12-24","title":"Analysis of Clinical Text: Task 14 of SemEval 2015","abstract":"SemEval (Semantic Evaluation) is an ongoing series of evaluations of computational semantic analysis systems, organized under the umbrella of SIGLEX, the Special Interest Group on the Lexicon of the Association for Computational Linguistics. This project describes Task 14 ("Analysis of Clinical Text") of the International Workshop on Semantic Evaluation 2015 (SemEval 2015). The purpose of Task 14 is to enhance current research in natural language processing (NLP) methods used in the clinical domain, and to introduce clinical text processing to the broader NLP community. The task aims to combine supervised methods for text analysis with unsupervised approaches. More specifically, the task aims to combine supervised methods for entity/acronym/abbreviation recognition and mapping to Unified Medical Language System (UMLS) Concept Unique Identifiers (CUIs) with access to larger clinical corpus for utilizing unsupervised techniques. Additionally, it will also evaluate systems on the task of template filling, which involves the population of eight attributes of the identified disorders with their normalized values.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":171204411,"compressed_storage_size":168769151,"resource_type":"Challenge","access_policy":"Credentialed","topics":["nlp","semeval"],"source_url":"https://physionet.org/content/semeval2015/1.0/"},{"public_project_uuid":"f4b9572a-c889-415c-9b53-50616a3b503e","slug":"challenge-2021","version":"1.0.0","core_doi":"10.13026/gt86-a263","version_doi":"10.13026/g9wd-3j66","is_latest_version":false,"short_description":"Will Two Do? Varying Dimensions in Electrocardiography: the PhysioNet/Computing in Cardiology Challenge 2021","publish_date":"2020-12-24","title":"Will Two Do? Varying Dimensions in Electrocardiography: The PhysioNet/Computing in Cardiology Challenge 2021","abstract":"The electrocardiogram (ECG) is a non-invasive representation of the electrical activity of the heart. Although the twelve-lead ECG is the standard diagnostic screening system for many cardiological issues, the limited accessibility of twelve-lead ECG devices provides a rationale for smaller, lower-cost, and easier to use devices. While single-lead ECGs are limiting, reduced-lead ECG systems hold promise, with evidence that subsets of the standard twelve leads can capture useful information and can even be comparable to twelve-lead ECGs in some limited contexts. In 2017 we challenged the public to classify AF from a single-lead ECG, and in 2020 we challenged the public to diagnose a much larger number of cardiac problems using twelve-lead recordings. However, there is limited evidence to demonstrate the utility of reduced-lead ECGs for capturing a wide range of diagnostic information.
\n\nIn this year's Challenge, we ask the following question: 'Will two do?' This year's Challenge builds on last year's Challenge, which asked participants to classify cardiac abnormalities from twelve-lead ECGs. We are asking you to build an algorithm that can classify cardiac abnormalities from either twelve-lead, six-lead, and two-lead ECGs. We will test each algorithm on databases of twelve-lead, six-lead and two-lead ECGs, and the differences in performances of the algorithms on these databases will reveal the utility of two-lead or six-lead ECGs in comparison to standard twelve-lead EGCs.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":14919,"compressed_storage_size":0,"resource_type":"Challenge","access_policy":"Open","topics":["challenge","cardiac abnormalities","competition","classification","multilead ecgs"],"source_url":"https://physionet.org/content/challenge-2021/1.0.0/"},{"public_project_uuid":"7f506139-c2ce-4518-b539-2f2e3d82c029","slug":"taichidb","version":"1.0.1","core_doi":"10.13026/9v35-xa84","version_doi":"10.13026/51h0-rs07","is_latest_version":false,"short_description":"This project includes gait data collected with footswitches and electromyography data from subjects who walked for 10 minutes under normal conditions and for 90 seconds under dual-task conditions (walking while performing serial subtractions).","publish_date":"2020-12-30","title":"Tai Chi, Physiological Complexity, and Healthy Aging: Gait and EMG data","abstract":"This dataset contains gait and electromyography (EMG) data collected during a hybrid study that included a two-arm randomized clinical trial (RCT) along with an additional observational comparison group. The RCT arm investigated the short-term effects of a Tai Chi intervention on 60 Tai Chi healthy naïve subjects, aged 50-79 years, living within the Greater Boston area, and reporting no regular Tai Chi practice within the past 5 years. Subjects were either randomized to 6 months of Tai Chi or usual care. Subjects came in for a baseline visit along with 3 month and 6 month follow-ups. The observational comparison group consisted of 27 healthy Tai Chi experts, aged 50-79 years, currently engaged in an active Tai Chi training regimen, each with at least 5 years of Tai Chi experience. The expert group was tested only at their baseline visit. The gait and EMG data were simultaneously recorded during walking at the subjects preferred speed for 10 minutes under the single-task condition and for 90 seconds under the dual-task condition. Our aim for this study was to use gait and EMG data to gain a better understanding of the mobility changes associated with Tai Chi training.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":2428074731,"compressed_storage_size":1596576421,"resource_type":"Database","access_policy":"Open","topics":["dual-task","stability","gait","electromyography"],"source_url":"https://physionet.org/content/taichidb/1.0.1/"},{"public_project_uuid":"7f1e44b7-f5a2-4470-8d4f-8af4be4355d2","slug":"semeval2015","version":"2.0","core_doi":"10.13026/96re-n451","version_doi":"10.13026/61rg-q298","is_latest_version":true,"short_description":"This is the dataset for SemEval 2014 and 2015, Analysis of Clinical Text","publish_date":"2014-12-28","title":"Analysis of Clinical Text: Task 14 of SemEval 2015","abstract":"SemEval (Semantic Evaluation) is an ongoing series of evaluations of computational semantic analysis systems, organized under the umbrella of SIGLEX, the Special Interest Group on the Lexicon of the Association for Computational Linguistics. This project describes "Analysis of Clinical Text" task of the International Workshop on Semantic Evaluation 2014 and 2015 (SemEval 2014 and 2015) [2,3,4,5]. The purpose of the task is to enhance current research in natural language processing (NLP) methods used in the clinical domain, and to introduce clinical text processing to the broader NLP community. The task aims to combine supervised methods for text analysis with unsupervised approaches for entity/acronym/abbreviation recognition and mapping to Unified Medical Language System (UMLS) Concept Unique Identifiers (CUIs). It also evaluated systems on the task of template filling [1], which involves the population of eight attributes of the identified disorders with their normalized values.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":3302893,"compressed_storage_size":3056703,"resource_type":"Challenge","access_policy":"Credentialed","topics":["nlp","semeval"],"source_url":"https://physionet.org/content/semeval2015/2.0/"},{"public_project_uuid":"5c11cbdc-7241-46b1-92d9-d7745c7c87f9","slug":"sufhsdb","version":"1.0.1","core_doi":"10.13026/9egj-jg49","version_doi":"10.13026/42eg-8e59","is_latest_version":true,"short_description":"The Shiraz University (SU) fetal heart sounds database (SUFHSDB) contains fetal and maternal phonocardiogram (PCG) recordings from 109 pregnant women in single and twin pregnancies.","publish_date":"2021-01-08","title":"Shiraz University Fetal Heart Sounds Database","abstract":"The Shiraz University (SU) fetal heart sounds database (SUFHSDB) contains fetal and maternal phonocardiogram (PCG) recordings from 109 pregnant women in single and twin pregnancies. The recordings were made at Hafez Hospital of Shiraz University of Medical Sciences, Shiraz, Iran, from mothers aged from 16 to 47 years with a digital stethoscope placed on the lower maternal abdomen. In total, 99 subjects had one signal recorded, three subjects had two and seven cases of twins were recorded individually, resulting in 119 recordings. The average duration of each record is about 90 seconds. The sampling rate was generally 16,000 Hz with 16-bit quantization and a few recordings at 44,100 Hz. The data was recorded in wide-band mode of the digital stethoscope, with a frequency response of 20 Hz to 1 kHz.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":827609336,"compressed_storage_size":663380299,"resource_type":"Database","access_policy":"Open","topics":["pregnancy","phonocardiography","fetal cardiography","cardiotocography"],"source_url":"https://physionet.org/content/sufhsdb/1.0.1/"},{"public_project_uuid":"eddbf60c-9d5a-4e68-bf9a-69ee4acde4d7","slug":"glucose-management-mimic","version":"1.0.0","core_doi":"10.13026/46zt-f692","version_doi":"10.13026/d18t-tf45","is_latest_version":false,"short_description":"The data subsets consist of time series files that includes all the curated entries of glucose readings and insulin inputs from MIMIC-III database.","publish_date":"2021-01-12","title":"Curated Data for Describing Blood Glucose Management in the Intensive Care Unit","abstract":"Analysis of real-world glucose and insulin clinical data recorded in electronic health records can provide insights into tailored approaches to clinical care, but still present many analytic challenges. The present data subsets are the result of a detailed curation process that extracts and pairs glucose readings to insulin therapy on a per-patient basis during an admission to the Intensive Care Unit (ICU) in the Medical Information Mart for Intensive Care (MIMIC-III) database. Curated data include over 500,000 glucose readings and more than 140,000 insulin entries for nearly 9,600 patients distributed in more than 11,000 ICU stays. Also, the proposed curation process involved the creation of glucose - insulin pairing rules according to clinical expert-defined physiologic and pharmacologic parameters. With the proposed rules, it was possible to pair nearly 76% of insulin administration events to a preceding blood glucose reading. The two shared data subsets have the potential to reveal insights regarding real-world practice for the glycemic control in the ICU. Moreover, the shared material serves as a framework for future studies of glucose management and insulin therapy replacement in the ICU, which may allow researchers to tailor queries and data processing to their own study objectives.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":137609127,"compressed_storage_size":15706708,"resource_type":"Database","access_policy":"Credentialed","topics":["insulin replacement therapy","critical care","glycemic control"],"source_url":"https://physionet.org/content/glucose-management-mimic/1.0.0/"},{"public_project_uuid":"0864bc47-d929-4a55-987c-756964a4bd6b","slug":"mimic-iii-question-answer","version":"1.0.0","core_doi":"10.13026/wrxb-6885","version_doi":"10.13026/j0y6-bw05","is_latest_version":true,"short_description":"Annotated Question Answering Pairs for Clinical Notes in the MIMIC-III Database","publish_date":"2021-01-15","title":"Annotated Question-Answer Pairs for Clinical Notes in the MIMIC-III Database","abstract":"Clinical question answering (QA) (or reading comprehension) aims to automatically answer questions from medical professionals based on clinical texts. We release this dataset, which contains 1287 annotated QA pairs on 36 sampled discharge summaries from MIMIC-III Clinical Notes, to facilitate the clinical question answering task. Questions in our dataset are either verified or directly generated by clinical experts.
\n\nNote that the primary purpose of this dataset is to test the generalizability of a QA model, i.e., whether a QA model that is trained on other datasets can answer questions on this dataset (which may have a different distribution compared with the training data), rather than to train a QA model. Hence the scale of our annotations is relatively small compared to some existing QA datasets.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":702703,"compressed_storage_size":176615,"resource_type":"Database","access_policy":"Credentialed","topics":["clinical question answering","clinical nlp","clinical reading comprehension"],"source_url":"https://physionet.org/content/mimic-iii-question-answer/1.0.0/"},{"public_project_uuid":"41d3d98c-89fe-4177-8124-be37f4a9019f","slug":"butppg","version":"1.0.0","core_doi":"10.13026/gxwp-p940","version_doi":"10.13026/7vy8-av04","is_latest_version":false,"short_description":"BUT PPG is a database created for the purpose of evaluating PPG signal quality and estimation of heart rate. The data comprises 48 10-second recordings of PPGs recorded by smartphone and associated ECG signals used for determination of reference HR.","publish_date":"2021-01-18","title":"Brno University of Technology Smartphone PPG Database (BUT PPG)","abstract":"Brno University of Technology Smartphone PPG Database (BUT PPG) is a database created by the cardiology team at the Department of Biomedical Engineering, Brno University of Technology, for the purpose of evaluating PPG quality and estimation of heart rate (HR). The data comprises 48 10-second recordings of PPGs and associated ECG signals used for determination of reference HR. The data were collected from 12 subjects (6 female, 6 male) aged between 21 to 61 years. Recordings were carried out between August 2020 and October 2020. PPG data were collected by smartphone Xiaomi Mi9 with sampling frequency of 30 Hz. Reference ECG signals were recorded using a mobile ECG recorder (Bittium Faros 360) with a sampling frequency of 1,000 Hz. Each PPG signal includes annotation of quality and reference HR. PPG signal quality is indicated binary: 1 indicates good quality for HR estimation, 0 indicates signals where HR cannot be detected reliably and thus these signals are unsuitable for any analysis.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":24703352,"compressed_storage_size":1296996,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/butppg/1.0.0/"},{"public_project_uuid":"49594a51-8bd8-41b3-841f-d64b4ea27713","slug":"but-pdb","version":"1.0.0","core_doi":"10.13026/br83-z194","version_doi":"10.13026/hwvj-5b53","is_latest_version":true,"short_description":"BUT PDB is an ECG signal database with marked peaks of P waves created for the development, and objective comparison of P wave detection algorithms. The database consists of 50 2-minute 2-lead ECG signal records with various types of pathology.","publish_date":"2021-01-19","title":"Brno University of Technology ECG Signal Database with Annotations of P Wave (BUT PDB)","abstract":"Brno University of Technology ECG Signal Database with Annotations of P Wave (BUT PDB) is an ECG signal database with marked peaks of P waves created by the cardiology team at the Department of Biomedical Engineering, Brno University of Technology. The database consists of 50 2-minute 2-lead ECG signal records with various types of pathology. The ECGs were selected from three existing databases of ECG signal - the MIT-BIH Arrhythmia Database, the MIT-BIH Supraventricular Arrhythmia Database, and the Long Term AF Database. The P waves positions were manually annotated by two ECG experts in all 50 records. Each record contains also annotation of positions and types of QRS complexes (from original database) and dominant diagnosis (pathology) present in record. This database is created for the development, evaluation and objective comparison of P wave detection algorithms.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":7373903,"compressed_storage_size":5161756,"resource_type":"Database","access_policy":"Open","topics":["ecg","p wave"],"source_url":"https://physionet.org/content/but-pdb/1.0.0/"},{"public_project_uuid":"0e4074b7-972f-4ffe-93e0-f6bb5b85a4c7","slug":"ludb","version":"1.0.1","core_doi":"10.13026/wdhq-rs83","version_doi":"10.13026/eegm-h675","is_latest_version":true,"short_description":"ECG signal database that consists of 200 10-second 12-lead records. The boundaries and peaks of P, T waves and QRS complexes were manually annotated by cardiologists. Each record is annotated with the corresponding diagnosis.","publish_date":"2021-01-19","title":"Lobachevsky University Electrocardiography Database","abstract":"Lobachevsky University Electrocardiography Database (LUDB) is an ECG signal database with marked boundaries and peaks of P, T waves and QRS complexes. The database consists of 200 10-second 12-lead ECG signal records representing different morphologies of the ECG signal. The ECGs were collected from healthy volunteers and patients of the Nizhny Novgorod City Hospital No 5 in 2017–2018. The patients had various cardiovascular diseases while some of them had pacemakers. The boundaries of P, T waves and QRS complexes were manually annotated by cardiologists for all 200 records. Also, each record is annotated with the corresponding diagnosis. The database can be used for educational purposes as well as for training and testing algorithms for ECG delineation, i.e. for automatic detection of boundaries and peaks of P, T waves and QRS complexes.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":24780747,"compressed_storage_size":14474484,"resource_type":"Database","access_policy":"Open","topics":["database","diagnosis","delineation","ecg","open database","electrocardiography"],"source_url":"https://physionet.org/content/ludb/1.0.1/"},{"public_project_uuid":"4da3e561-e167-41ff-9815-c47c56d62455","slug":"cerebral-perfusion-diabetes","version":"1.0.0","core_doi":"10.13026/pgn6-6k53","version_doi":"10.13026/rbeh-9r20","is_latest_version":false,"short_description":"Dataset collected during a study on type 2 diabetes on brain blood flow, vasoreactivity and functional outcomes (gait and balance) using TCD, MRI perfusion and foot pressure distribution and gait measures.","publish_date":"2021-01-20","title":"Cerebral perfusion and cognitive decline in type 2 diabetes","abstract":"This dataset was collected as part of a study to explore vasoregulation and blood flow in patients with type 2 Diabetes mellitus. It comprises cerebral blood, autonomic functions, laboratory and gait variables in 70 patients with type 2 DM and 70 healthy controls (50-85 years old). Diabetes mellitus is a prevalent, costly condition with high morbidity and mortality that affects nearly 6.3% of the population in the United States. Diabetes mellitus alters the permeability of the blood-brain barrier, thus affecting regional metabolism and microcirculatory regulation. Predictors of cerebrovascular complications of diabetes that are evidence-based on cerebral blood flow, MRI imaging, and cognitive testing are lacking.
\n\nBlood flow velocities were measured in the anterior and middle cerebral arteries using transcranial Doppler ultrasound. Cerebral vasoregulation was evaluated by comparing blood flow velocity responses during hypocapnia and hypercapnia, Valsalva maneuver, head-up tilt and sit-to-stand test using simultaneous recordings of cardiovascular variables, blood flow velocity in the anterior and middle cerebral artery using transcranial Doppler and respiratory variables. We applied continuous arterial spin labeling at 3 Tesla MRI to evaluate regional distribution of cerebral blood flow and vasomotor reactivity to CO2. Cognitive and executive functions were assessed via a battery of psychometric measures. Gait and balance were characterized from dynamics of foot pressure distribution and center of pressure displacement.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":13990755206,"compressed_storage_size":5435954857,"resource_type":"Database","access_policy":"Open","topics":["vasoregulation","diabetes","brain"],"source_url":"https://physionet.org/content/cerebral-perfusion-diabetes/1.0.0/"},{"public_project_uuid":"b0a620eb-9e55-4762-afd5-3a2c98b3f488","slug":"stroke-scale-mimic-iii","version":"1.0.0","core_doi":"10.13026/jb5v-w976","version_doi":"10.13026/gyjg-0t90","is_latest_version":true,"short_description":"A dataset of annotated NIHSS scale items and corresponding scores from stroke patients discharge summaries in MIMIC-III.","publish_date":"2021-01-25","title":"National Institutes of Health Stroke Scale (NIHSS) Annotations for the MIMIC-III Database","abstract":"The National Institutes of Health Stroke Scale (NIHSS) is a 15-item neurologic examination stroke scale. It quantifies the physical manifestations of neurological deficits and provides crucial support for clinical decision making and early-stage emergency triage. NIHSS scores stored in the free-text of Electronic Health Records (EHRs) often lack standardization and the expression patterns are highly dependent on the habit of the clinicians. This can limit the potential for reusability of the data.
\n\nThere is benefit in developing robust algorithms to extract NIHSS scores from the free-text of EHRs. We developed a dataset for NIHSS score identification, a task defined as the extraction of scale items and corresponding scores from discharge summaries. Discharge summaries of stroke cases in the Medical Information Mart for Intensive Care III (MIMIC-III) database were used to create an annotated NIHSS corpus.
\n\nEach discharge summary was manually annotated for the presence of NIHSS scores by two annotators with backgrounds in medical informatics. Annotations include all scale items (e.g. “4. Facial Palsy”), the corresponding score “measurement”, and their relation “has value”. The dataset is intended to support academic and industrial research in the field of medical natural language processing (NLP).
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1708168,"compressed_storage_size":221417,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/stroke-scale-mimic-iii/1.0.0/"},{"public_project_uuid":"5c4e3f33-b2ab-4987-b1c2-75782fcb25d3","slug":"body-sway-music-vr","version":"1.0.0","core_doi":"10.13026/76ht-2717","version_doi":"10.13026/x32c-cz47","is_latest_version":true,"short_description":"This data were intended to show that music manipulated to match VR motion provided by an Oculus Rift head mounted display increased body sway when standing still.","publish_date":"2021-01-26","title":"Body Sway When Standing and Listening to Music Modified to Reinforce Virtual Reality Environment Motion","abstract":"The increased likelihood of falls as a consequence of aging or disease is generally associated with rising levels of body sway during stance. The use of virtual reality (VR) for studying the role in vision in regulating body sway is well established, however the extent to which music incorporated into a virtual environment can influence sway is not understood. This dataset was collected as part of a study to explore the hypothesis that music manipulated to match VR motion provided by an Oculus Rift head mounted display can lead to increased levels of body sway. Twenty-eight subjects stood for 60 s on a balance platform that measured anterior posterior (AP) and medial lateral (ML) center of pressure movement (indicators of body sway). While standing, the subjects experienced combinations of 3 visual conditions (VR translation in the AP direction at 0.1 Hz, no translation, and eyes closed) and 4 music conditions (Mozart’s Jupiter Symphony modified to scale volume at 0.1 Hz and 0.25 Hz, unmodified music, and no music) for a total of 12 trials. Analysis of the data using frequency domain measures may reveal the extent to which music influenced body sway.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1208953437,"compressed_storage_size":277039002,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/body-sway-music-vr/1.0.0/"},{"public_project_uuid":"75edd3b4-3655-4faf-85e0-f7bf6d5c3e55","slug":"mimic-cxr-pe-severity","version":"1.0","core_doi":"10.13026/e4ej-gm40","version_doi":"10.13026/qy76-9178","is_latest_version":false,"short_description":"Pulmonary edema metadata and labels for MIMIC-CXR","publish_date":"2021-01-26","title":"Pulmonary Edema Severity Grades Based on MIMIC-CXR","abstract":"Clinical management decisions for patients with acutely decompensated heart failure and many other diseases are often based on grades of pulmonary edema severity, rather than its mere absence or presence. Chest radiographs are commonly performed to assess pulmonary edema. The MIMIC-CXR dataset that consists of 377,110 chest radiographs with free-text radiology reports offers a tremendous opportunity to study this subject.
\n\nThis dataset is curated based on MIMIC-CXR, containing 3 metadata files that consist of pulmonary edema severity grades extracted from the MIMIC-CXR dataset through different means: 1) by regular expression (regex) from radiology reports, 2) by expert labeling from radiology reports, and 3) by consensus labeling from chest radiographs.
\n\nThis dataset aims to support the algorithmic development of pulmonary edema assessment from chest x-ray images and benchmark its performance. The metadata files have subject IDs, study IDs, DICOM IDs, and the numerical grades of pulmonary edema severity. The IDs listed in this dataset have the same mapping structure as in MIMIC-CXR.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":33662138,"compressed_storage_size":12663893,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/mimic-cxr-pe-severity/1.0/"},{"public_project_uuid":"c7331495-c03a-429b-b033-99133f99938a","slug":"challenge-2021","version":"1.0.1","core_doi":"10.13026/gt86-a263","version_doi":"10.13026/13he-1932","is_latest_version":false,"short_description":"Will Two Do? Varying Dimensions in Electrocardiography: the PhysioNet/Computing in Cardiology Challenge 2021","publish_date":"2021-02-01","title":"Will Two Do? Varying Dimensions in Electrocardiography: The PhysioNet/Computing in Cardiology Challenge 2021","abstract":"The electrocardiogram (ECG) is a non-invasive representation of the electrical activity of the heart. Although the twelve-lead ECG is the standard diagnostic screening system for many cardiological issues, the limited accessibility of twelve-lead ECG devices provides a rationale for smaller, lower-cost, and easier to use devices. While single-lead ECGs are limiting, reduced-lead ECG systems hold promise, with evidence that subsets of the standard twelve leads can capture useful information and can even be comparable to twelve-lead ECGs in some limited contexts. In 2017 we challenged the public to classify AF from a single-lead ECG, and in 2020 we challenged the public to diagnose a much larger number of cardiac problems using twelve-lead recordings. However, there is limited evidence to demonstrate the utility of reduced-lead ECGs for capturing a wide range of diagnostic information.
\n\nIn this year's Challenge, we ask the following question: 'Will two do?' This year's Challenge builds on last year's Challenge, which asked participants to classify cardiac abnormalities from twelve-lead ECGs. We are asking you to build an algorithm that can classify cardiac abnormalities from either twelve-lead, six-lead, three-lead, and two-lead ECGs. We will test each algorithm on databases of twelve-lead, six-lead, three-lead, and two-lead ECGs, and the differences in performances of the algorithms on these databases will reveal the utility of reduced-lead ECGs in comparison to standard twelve-lead EGCs.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":14919,"compressed_storage_size":0,"resource_type":"Challenge","access_policy":"Open","topics":["classification","competition","cardiac abnormalities","challenge","multilead ecgs"],"source_url":"https://physionet.org/content/challenge-2021/1.0.1/"},{"public_project_uuid":"ba83b7d9-8145-44d0-968a-92c57a9c0462","slug":"hirid","version":"1.1","core_doi":"10.13026/323r-nk04","version_doi":"10.13026/mn2d-nq25","is_latest_version":false,"short_description":"The HiRID database contains a large selection of all routinely collected data relating to patient admissions to the Department of Intensive Care Medicine of the Bern University Hospital, Switzerland (ICU).","publish_date":"2021-02-04","title":"HiRID, a high time-resolution ICU dataset","abstract":"HiRID is a freely accessible critical care dataset containing data relating to almost 34 thousand patient admissions to the Department of Intensive Care Medicine of the Bern University Hospital, Switzerland (ICU), an interdisciplinary 60-bed unit admitting >6,500 patients per year. The ICU offers the full range of modern interdisciplinary intensive care medicine for adult patients. The dataset was developed in cooperation between the Swiss Federal Institute of Technology (ETH) Zürich, Switzerland and the ICU.
\n\nThe dataset contains de-identified demographic information and a total of 681 routinely collected physiological variables, diagnostic test results and treatment parameters from almost 34 thousand admissions during the period from January 2008 to June 2016. Data is stored with a uniquely high time resolution of one entry every two minutes.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":18024865631,"compressed_storage_size":17996737657,"resource_type":"Database","access_policy":"Contributor Review","topics":["critical care","clinical","intensive care","icu","high resolution","machine learning"],"source_url":"https://physionet.org/content/hirid/1.1/"},{"public_project_uuid":"ef340a80-898d-4072-a6d6-cbf027629e6b","slug":"mimic-cxr-pe-severity","version":"1.0.1","core_doi":"10.13026/e4ej-gm40","version_doi":"10.13026/rz5p-rc64","is_latest_version":true,"short_description":"Pulmonary edema metadata and labels for MIMIC-CXR","publish_date":"2021-02-09","title":"Pulmonary Edema Severity Grades Based on MIMIC-CXR","abstract":"Clinical management decisions for patients with acutely decompensated heart failure and many other diseases are often based on grades of pulmonary edema severity, rather than its mere absence or presence. Chest radiographs are commonly performed to assess pulmonary edema. The MIMIC-CXR dataset that consists of 377,110 chest radiographs with free-text radiology reports offers a tremendous opportunity to study this subject.
\n\nThis dataset is curated based on MIMIC-CXR, containing 3 metadata files that consist of pulmonary edema severity grades extracted from the MIMIC-CXR dataset through different means: 1) by regular expression (regex) from radiology reports, 2) by expert labeling from radiology reports, and 3) by consensus labeling from chest radiographs.
\n\nThis dataset aims to support the algorithmic development of pulmonary edema assessment from chest x-ray images and benchmark its performance. The metadata files have subject IDs, study IDs, DICOM IDs, and the numerical grades of pulmonary edema severity. The IDs listed in this dataset have the same mapping structure as in MIMIC-CXR.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":33662138,"compressed_storage_size":12663917,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/mimic-cxr-pe-severity/1.0.1/"},{"public_project_uuid":"24139bfa-6a5f-491f-b285-a73ac019e7f9","slug":"bigideaslab-step-hr-smartwatch","version":"1.0","core_doi":"10.13026/z82k-3w96","version_doi":"10.13026/cqfy-d860","is_latest_version":true,"short_description":"Comparison of HR values reported by ECG (Bittium Faros) and wearables including Biovotion Everion, Empatica E4, Apple Watch, Garmin, Fitbit, and Xiaomi Miband.","publish_date":"2021-02-10","title":"BigIdeasLab_STEP: Heart rate measurements captured by smartwatches for differing skin tones","abstract":"This dataset was collected as part of a study on investigating the sources of inaccuracy in wearable optical heart rate sensors across a variety of activities (sitting, deep breathing, walking, and typing). The original study examined covariates including skin tone, signal lag, and device type. The dataset includes the heart rate given by the ECG and corresponding heart rate given by each smartwatch used in the study. Activity, skin tone, and subject ID are also included. This dataset was recorded in July-August 2019 and includes 53 participants. The data is completely de-identified and observations are time-synced.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":12270238,"compressed_storage_size":1532377,"resource_type":"Database","access_policy":"Restricted","topics":["wearables","heart rate"],"source_url":"https://physionet.org/content/bigideaslab-step-hr-smartwatch/1.0/"},{"public_project_uuid":"d27bfcb8-158f-4c10-8b4c-0cb45246b695","slug":"hirid","version":"1.1.1","core_doi":"10.13026/323r-nk04","version_doi":"10.13026/nkwc-js72","is_latest_version":true,"short_description":"The HiRID database contains a large selection of all routinely collected data relating to patient admissions to the Department of Intensive Care Medicine of the Bern University Hospital, Switzerland (ICU).","publish_date":"2021-02-18","title":"HiRID, a high time-resolution ICU dataset","abstract":"HiRID is a freely accessible critical care dataset containing data relating to almost 34 thousand patient admissions to the Department of Intensive Care Medicine of the Bern University Hospital, Switzerland (ICU), an interdisciplinary 60-bed unit admitting >6,500 patients per year. The ICU offers the full range of modern interdisciplinary intensive care medicine for adult patients. The dataset was developed in cooperation between the Swiss Federal Institute of Technology (ETH) Zürich, Switzerland and the ICU.
\n\nThe dataset contains de-identified demographic information and a total of 681 routinely collected physiological variables, diagnostic test results and treatment parameters from almost 34 thousand admissions during the period from January 2008 to June 2016. Data is stored with a uniquely high time resolution of one entry every two minutes.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":18024882962,"compressed_storage_size":17996760489,"resource_type":"Database","access_policy":"Contributor Review","topics":["critical care","high resolution","intensive care","icu","clinical","machine learning"],"source_url":"https://physionet.org/content/hirid/1.1.1/"},{"public_project_uuid":"6a4fb7e1-4029-432c-a556-bb30af4ae6b7","slug":"challenge-2021","version":"1.0.2","core_doi":"10.13026/gt86-a263","version_doi":"10.13026/jz9p-0m02","is_latest_version":false,"short_description":"Will Two Do? Varying Dimensions in Electrocardiography: the PhysioNet/Computing in Cardiology Challenge 2021","publish_date":"2021-02-25","title":"Will Two Do? Varying Dimensions in Electrocardiography: The PhysioNet/Computing in Cardiology Challenge 2021","abstract":"The electrocardiogram (ECG) is a non-invasive representation of the electrical activity of the heart. Although the twelve-lead ECG is the standard diagnostic screening system for many cardiological issues, the limited accessibility of twelve-lead ECG devices provides a rationale for smaller, lower-cost, and easier to use devices. While single-lead ECGs are limiting [1], reduced-lead ECG systems hold promise, with evidence that subsets of the standard twelve leads can capture useful information [2], [3], [4] and even be comparable to twelve-lead ECGs in some limited contexts. In 2017 we challenged the public to classify AF from a single-lead ECG, and in 2020 we challenged the public to diagnose a much larger number of cardiac problems using twelve-lead recordings. However, there is limited evidence to demonstrate the utility of reduced-lead ECGs for capturing a wide range of diagnostic information.
\n\nIn this year’s Challenge, we ask the following question: ‘Will two do?’ This year’s Challenge builds on last year’s Challenge [5], which asked participants to classify cardiac abnormalities from twelve-lead ECGs. We are asking you to build an algorithm that can classify cardiac abnormalities from twelve-lead, six-lead, four-lead, three-lead, and two-lead ECGs. We will test each algorithm on databases of these reduced-lead ECGs, and the differences in performances of the algorithms on these databases will reveal the utility of reduced-lead ECGs in comparison to standard twelve-lead EGCs.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":14919,"compressed_storage_size":0,"resource_type":"Challenge","access_policy":"Open","topics":["classification","challenge","multilead ecgs","competition","cardiac abnormalities"],"source_url":"https://physionet.org/content/challenge-2021/1.0.2/"},{"public_project_uuid":"1e4f25d1-e5ce-46dc-8891-e7b215b804df","slug":"flatten-covid-survey","version":"1.0","core_doi":"10.13026/3pqw-8411","version_doi":"10.13026/v8eq-8v80","is_latest_version":true,"short_description":"Freely accessible COVID-19 symptom dataset surveying Canadians and gathered from March to July of 2020 by the global humanitarian aid non-profit Flatten. This dataset of 294,106 surveys gathered from March 23rd to July 30th in 2020.","publish_date":"2021-03-08","title":"Flatten: COVID-19 Survey Data on Symptoms, Demographics and Mental Health in Canada","abstract":"The expedited collection and analysis of pre-clinical syndromic data is delivering novel insights into public health conditions during the ongoing COVID-19 pandemic. We provide Canada’s first publicly available pre-clinical COVID-19 dataset, based on survey responses collected from 294,106 Canadians from March 23rd until July 30th 2020, using a platform developed by Flatten, a Canadian non-profit organization. We provide this data for academic and industry research.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":9615386,"compressed_storage_size":979976,"resource_type":"Database","access_policy":"Restricted","topics":["covid-19","public health","population statistics"],"source_url":"https://physionet.org/content/flatten-covid-survey/1.0/"},{"public_project_uuid":"bce4cce1-073e-48dc-af41-828065a265e1","slug":"mimiciv","version":"1.0","core_doi":"10.13026/07hj-2a80","version_doi":"10.13026/s6n6-xd98","is_latest_version":false,"short_description":"Database of patients admitted to the Beth Israel Deaconess Medical Center","publish_date":"2021-03-16","title":"MIMIC-IV","abstract":"Retrospectively collected medical data has the opportunity to improve patient care through knowledge discovery and algorithm development. Broad reuse of medical data is desirable for the greatest public good, but data sharing must be done in a manner which protects patient privacy. The Medical Information Mart for Intensive Care (MIMIC)-III database provided critical care data for over 40,000 patients admitted to intensive care units at the Beth Israel Deaconess Medical Center (BIDMC). Importantly, MIMIC-III was deidentified, and patient identifiers were removed according to the Health Insurance Portability and Accountability Act (HIPAA) Safe Harbor provision. MIMIC-III has been integral in driving large amounts of research in clinical informatics, epidemiology, and machine learning. Here we present MIMIC-IV, an update to MIMIC-III, which incorporates contemporary data and improves on numerous aspects of MIMIC-III. MIMIC-IV adopts a modular approach to data organization, highlighting data provenance and facilitating both individual and combined use of disparate data sources. MIMIC-IV is intended to carry on the success of MIMIC-III and support a broad set of applications within healthcare.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":7441125832,"compressed_storage_size":7382623686,"resource_type":"Database","access_policy":"Credentialed","topics":["mimic","intensive care unit","machine learning","critical care"],"source_url":"https://physionet.org/content/mimiciv/1.0/"},{"public_project_uuid":"63ff0df3-ab70-4225-b2a7-bc399f3bc687","slug":"wfdb","version":"10.6.2","core_doi":"10.13026/zx10-se90","version_doi":"10.13026/zzpx-h016","is_latest_version":false,"short_description":"The original Waveform Database Software (WFDB) Package, in C.","publish_date":"2021-03-26","title":"WFDB Software Package","abstract":"Effective processing and analysis of physiological data requires specialized software. We have developed a large collection of such software over the past thirty years, and much of it is contained within the WFDB (Waveform Database) Software Package. The WFDB Software Package comprises over 70 applications for signal processing and automated analysis. A comprehensive collection of documentation, including tutorials and reference manuals, is also included in the package. The package is written in highly portable C and can be used on all popular platforms, including GNU/Linux, MacOS, MS-Windows, and all versions of Unix. The package has also been ported to other popular languages, including Matlab and Python.
","license":{"name":"GNU General Public License version 3"},"dua":null,"main_storage_size":7248271,"compressed_storage_size":3921088,"resource_type":"Software","access_policy":"Open","topics":["wfdb","waveform"],"source_url":"https://physionet.org/content/wfdb/10.6.2/"},{"public_project_uuid":"3610ce94-a384-4ec5-9779-c38bea463a42","slug":"wfdb-python","version":"3.3.0","core_doi":"10.13026/jt5e-gq15","version_doi":"10.13026/g35g-c061","is_latest_version":false,"short_description":"Tools for working with waveforms in Python.","publish_date":"2021-03-26","title":"Waveform Database Software Package (WFDB) for Python","abstract":"Physiological waveforms - such as electrocardiograms (ECG), electroencephalograms (EEG), electromyograms (EMG) - are generated during the course of routine care. These signals contain information that can be used to understand underlying conditions of health. Effective processing and analysis of physiological data requires specialized software. The waveform-database (WFDB) package for Python is a library of tools for reading, writing, and processing physiological signals and annotations, written in the Python programming language. Core components of the Python package are loosely based on the specifications of the original C-language WFDB software. We aim to implement as many of the core WFDB features as possible, with user-friendly APIs. Additional features are added over time.
","license":{"name":"MIT License"},"dua":null,"main_storage_size":333684020,"compressed_storage_size":100662741,"resource_type":"Software","access_policy":"Open","topics":["wfdb","waveform","python"],"source_url":"https://physionet.org/content/wfdb-python/3.3.0/"},{"public_project_uuid":"bfdb0e6f-2de8-407c-9202-ad7fa7d397ec","slug":"wfdb-matlab","version":"0.10.0","core_doi":"10.13026/yry5-ds12","version_doi":"10.13026/6zcz-e163","is_latest_version":true,"short_description":"Tools for working with waveforms in MATLAB.","publish_date":"2021-04-05","title":"Waveform Database Software Package (WFDB) for MATLAB and Octave","abstract":"Physiological waveforms - such as electrocardiograms (ECG), electroencephalograms (EEG), electromyograms (EMG) - are generated during the course of routine care. These signals contain information that can be used to understand underlying conditions of health. Effective processing and analysis of physiological data requires specialized software. The WaveForm DataBase (WFDB) Toolbox for MATLAB and Octave is a collection of over 30 functions and utilities that integrate PhysioNet's open-source applications and databases with the high-precision numerical computational and graphics environment of MATLAB and Octave.
","license":{"name":"GNU General Public License version 3"},"dua":null,"main_storage_size":8333133,"compressed_storage_size":4642307,"resource_type":"Software","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/wfdb-matlab/0.10.0/"},{"public_project_uuid":"03b1dd30-71e0-4438-95d0-e7894654cb42","slug":"eeg-power-anesthesia","version":"1.0.0","core_doi":"10.13026/bnj3-7b52","version_doi":"10.13026/m792-h077","is_latest_version":true,"short_description":"EEG power spectra recorded during anesthesia","publish_date":"2021-04-19","title":"Multitaper spectra recorded during GABAergic anesthetic unconsciousness","abstract":"This database contains electroencephalography (EEG) multitaper spectra and associated conscious/unconscious labels for a cohort of 10 healthy volunteers undergoing stereotyped anesthetic administration and direct monitoring of subject responsiveness, and 44 patients receiving anesthesia care in an operating room (OR) context. Cases recorded in an OR context are further divided into those administered propofol alone under a total intravenous anesthesia (TIVA) approach, and those who received either sevoflurane alone or sevoflurane following a propofol induction. These data were used to train and validate classification models for tracking unconsciousness. Corresponding code for building and testing classification models is publicly available.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":860175967,"compressed_storage_size":263801454,"resource_type":"Database","access_policy":"Restricted","topics":["electroencephalography","eeg","unconsciousness","anesthesia"],"source_url":"https://physionet.org/content/eeg-power-anesthesia/1.0.0/"},{"public_project_uuid":"e34a1458-1048-4c7f-96cc-dfac75d345b3","slug":"glucose-management-mimic","version":"1.0.1","core_doi":"10.13026/46zt-f692","version_doi":"10.13026/517s-2q57","is_latest_version":true,"short_description":"The data subsets consist of time series files that includes all the curated entries of glucose readings and insulin inputs from MIMIC-III database.","publish_date":"2021-04-19","title":"Curated Data for Describing Blood Glucose Management in the Intensive Care Unit","abstract":"Analysis of real-world glucose and insulin clinical data recorded in electronic health records can provide insights into tailored approaches to clinical care, but still present many analytic challenges. The present data subsets are the result of a detailed curation process that extracts and pairs glucose readings to insulin therapy on a per-patient basis during an admission to the Intensive Care Unit (ICU) in the Medical Information Mart for Intensive Care (MIMIC-III) database. Curated data include over 500,000 glucose readings and more than 140,000 insulin entries for nearly 9,600 patients distributed in more than 11,000 ICU stays. Also, the proposed curation process involved the creation of glucose - insulin pairing rules according to clinical expert-defined physiologic and pharmacologic parameters. With the proposed rules, it was possible to pair nearly 76% of insulin administration events to a preceding blood glucose reading. The two shared data subsets have the potential to reveal insights regarding real-world practice for the glycemic control in the ICU. Moreover, the shared material serves as a framework for future studies of glucose management and insulin therapy replacement in the ICU, which may allow researchers to tailor queries and data processing to their own study objectives.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":137609127,"compressed_storage_size":15706708,"resource_type":"Database","access_policy":"Credentialed","topics":["insulin replacement therapy","glycemic control","critical care"],"source_url":"https://physionet.org/content/glucose-management-mimic/1.0.1/"},{"public_project_uuid":"7bb26ab6-3125-4715-af0a-d3740b858001","slug":"mental-fnirs","version":"1.0","core_doi":"10.13026/m32v-hy93","version_doi":"10.13026/zfb2-1g43","is_latest_version":true,"short_description":"Functional near-infrared spectroscopy and transcranial doppler sonography were used to measure changes in cerebral hemodynamics during cognitive stimulation.","publish_date":"2021-04-21","title":"Mental workload during n-back task captured by TransCranial Doppler (TCD) sonography and functional Near-Infrared Spectroscopy (fNIRS) monitoring","abstract":"The authors used this dataset to assess the neurovascular coupling responses evoked by a cognitive paradigm with the aid of TransCranial Doppler (TCD) sonography and functional Near-InfraRed Spectroscopy (fNIRS). Fourteen healthy young individuals were recruited for this study conducted in 2018 at the University of Oklahoma, Translational Geroscience Laboratory (age: 31±5.94 years, BMI: 24.9±2.95). The following physiological signals were measured non-invasively during the examination: relative concentration of oxy- and deoxyhemoglobin (HbO and HbR, respectively), and cerebral Blood Flow Velocity (BFV) in the middle cerebral arteries for the entire duration of the testing (~11 minutes). In this study, fNIRS signals from the frontal cortex were acquired by a NIRScout platform (NIRx Medical Technologies LLC, NY, USA) at a sampling frequency of 3.9 Hz, HbO and HbR time series signals can be obtained by applying the differential modified Beer Lambert Law on the preprocessed records of near-infrared intensities. BFV was recorded using TCD sonography (Digi-Lite, Rimed, Raanana, Israel) by placing 2 Mhz ultrasound probes over the left and right temporal acoustic windows. The paradigm used for cognitive testing was an n-back task where the participant was administered a series of alphabetical letters on the computer screen and had to provide a response whenever a stimulus is presented that is the same as the one presented n trials previously. The examination protocol consisted of the following 134 second sessions: 0-back, 1-back, 0-back, and 2-back. This dataset was collected for the following study: "Increased Cognitive Workload Evokes Greater Neurovascular Coupling Responses in Healthy Young Adults".
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":242183461,"compressed_storage_size":102080874,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/mental-fnirs/1.0/"},{"public_project_uuid":"ec2c108f-ff34-4d4f-a94a-17e40623bef4","slug":"sleep-dep-hemo-cog","version":"1.0.0","core_doi":"10.13026/jv2f-az31","version_doi":"10.13026/669h-cx11","is_latest_version":true,"short_description":"We assessed the impact of 24-hour sleep deprivation on the global properties of frontal lobe functional networks and on cognitive performance. fNIRS measurements were carried out in the resting state and in response to a finger-tapping paradigm.","publish_date":"2021-04-26","title":"Effect of 24-hour sleep deprivation on cerebral hemodynamics and cognitive performance","abstract":"Sleep deprivation (SD) is associated with decreased cognitive performance, but the underlying mechanisms are poorly understood. To assess the impact of SD on cerebral hemodynamics in the frontal lobe, functional near-infrared spectroscopy (fNIRS) measurements were carried out on young subjects. Ten young healthy adults (males, age of 27.6±3.7 years, one left-handed participant) were recruited for this study conducted in 2018 at University of Oklahoma, Translational Geroscience Laboratory. The examination protocol began with assessment of cognitive performance with the aid of Cambridge Neuropsychological Test Automated Battery (CANTAB). The following tests were used: motor screening task, rapid visual processing, reaction time, spatial working memory, paired associated learning, and delayed matching to visual sample. Subsequently fNIRS measurements were taken from participants while carrying out a finger-tapping exercise. Each subject was presented with an auditory command to perform a finger-tapping task over three sets of 10 seconds using the left index finger and three sets of 10 seconds using the right index finger, with 10 seconds inter-stimulus periods between each task. Functional NIRS recordings were acquired using a NIRScout platform (NIRx Medical Technologies LLC, NY, USA). Relative oxy- and deoxyhemoglobin time series can be obtained from measured intensities by applying differential modified Beer Lamber Law. All measurements were performed before and after 24 hours of SD. The dataset forms the basis for two studies that explore: (1) the relationship between sleep deprivation, cognitive performance, and neurovascular coupling response in young healthy adults; and (2) sleep deprivation and functional connectivity of the frontal cortex.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":162066739,"compressed_storage_size":89191320,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/sleep-dep-hemo-cog/1.0.0/"},{"public_project_uuid":"20f97372-6da4-41c7-96f0-e06b2abbb669","slug":"clinical-bert-mimic-notes","version":"1.0.0","core_doi":"10.13026/e4hr-yh29","version_doi":"10.13026/vk2z-bz63","is_latest_version":true,"short_description":"We explore recovering sensitive info from BERT trained over non-deidentified EHR. We make our models and data available to further facilitate research.","publish_date":"2021-04-28","title":"Clinical BERT Models Trained on Pseudo Re-identified MIMIC-III Notes","abstract":"This project contains the weights of 7 different BERT models trained over a surrogate re-identified set of MIMIC-III (v1.4) notes. Each model is suitable for direct compatibility with the HuggingFace framework; users can easily load and apply the model. We release these model weights with the intent of facilitating research into the dangers of revealing patient information via model sharing, specifically for models pretrained on non-deidentified electronic health record (EHR) data. We additionally release our post-processed data, which most importantly contains a mapping of patient subject id to condition(s) that the given patient has. The purpose of these "subject id to condition" files is to measure how much of a correlation can be extracted from a surrogate patient name and the medical conditions that the given patient has. This will allow users to quantify the amount of Protected Health Information (PHI) "leakage" these large pretrained language models may exhibit.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":9474970963,"compressed_storage_size":5604946282,"resource_type":"Model","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/clinical-bert-mimic-notes/1.0.0/"},{"public_project_uuid":"fa1dd19e-f8c6-4c0d-b950-cfb4f5c540d4","slug":"mimic-eicu-fiddle-feature","version":"1.0.0","core_doi":"10.13026/zwds-nn76","version_doi":"10.13026/2qtg-k467","is_latest_version":true,"short_description":"Features and labels from MIMIC-III and eICU-CRD produced by FIDDLE, an EHR preprocessing pipeline.","publish_date":"2021-04-29","title":"MIMIC-III and eICU-CRD: Feature Representation by FIDDLE Preprocessing","abstract":"This is a preprocessed dataset derived from patient records in MIMIC-III and eICU, two large-scale electronic health record (EHR) databases. It contains features and labels for 5 prediction tasks involving 3 adverse outcomes (prediction times listed in parentheses): in-hospital mortality (48h), acute respiratory failure (4h and 12h), and shock (4h and 12h). We extracted comprehensive, high-dimensional feature representations (up to ~8,000 features) using FIDDLE (FlexIble Data-Driven pipeLinE), an open-source preprocessing pipeline for structured clinical data. These 5 prediction tasks were designed in consultation with a critical care physician for their clinical importance, and were used as part of the proof-of-concept experiments in the original paper to demonstrate FIDDLE's utility in aiding the feature engineering step of machine learning model development. The intent of this release is to share preprocessed MIMIC-III and eICU datasets used in the experiments to support and enable reproducible machine learning research on EHR data.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1794482800,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":["electronic health record","preprocessing","machine learning"],"source_url":"https://physionet.org/content/mimic-eicu-fiddle-feature/1.0.0/"},{"public_project_uuid":"ab0d2501-4390-472b-a818-e30e3373ac17","slug":"treadmill-exercise-cardioresp","version":"1.0.0","core_doi":"10.13026/94yv-6145","version_doi":"10.13026/de8x-qg95","is_latest_version":false,"short_description":"Cardiorespiratory measurements of 992 treadmill maximal graded exercise tests. Heart rate, oxygen consumption, carbon dioxide generation, and pulmonary ventilation are provided.","publish_date":"2021-04-30","title":"Treadmill Maximal Exercise Tests from the Exercise Physiology and Human Performance Lab of the University of Malaga","abstract":"The present database is an ensemble of the cardiorespiratory measurements acquired during 992 treadmill maximal graded exercise tests (GET) performed in the Exercise Physiology and Human Performance Lab of the University of Malaga. Heart rate, oxygen consumption, carbon dioxide generation, and pulmonary ventilation are measured on a breath-to-breath basis along with the treadmill speed during maximal effort tests. Participants are amateur and professional athletes are of ages ranging from 10 to 63 years old. The age, height, and weight of the participants are provided, as well as the temperature and humidity during the test.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":23193152,"compressed_storage_size":6572835,"resource_type":"Database","access_policy":"Open","topics":["heart rate","cardiovascular fitness","heart rate recovery","oxygen consumption","graded exercise testing","cardiorespiratory measures","exercise testing","fitness monitoring"],"source_url":"https://physionet.org/content/treadmill-exercise-cardioresp/1.0.0/"},{"public_project_uuid":"6e2940f0-e209-4bd8-9e37-172748cf86f2","slug":"eicu-crd-demo","version":"2.0.1","core_doi":"10.13026/krfm-4167","version_doi":"10.13026/4mxk-na84","is_latest_version":true,"short_description":"An openly available subset of the eICU Collaborative Research Database.","publish_date":"2021-05-07","title":"eICU Collaborative Research Database Demo","abstract":"This project is a demo version of the eICU Collaborative Research Database, a multi-center database comprised of deidentified health data for over 200,000 admissions to ICUs across the United States between 2014-2015. The database includes vital sign measurements, care plan documentation, severity of illness measures, diagnosis information, and treatment information. Data is collected through the Philips eICU program, a critical care telehealth program that delivers information to caregivers at the bedside. This open access demo allows researchers to ascertain whether the eICU Collaborative Research Database is suitable for their work. It includes over 2,500 unit stays selected from 20 of the larger hospitals in the eICU Collaborative Research Database.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":136979508,"compressed_storage_size":136773541,"resource_type":"Database","access_policy":"Open","topics":["critical care","medical informatics","open data","eicu"],"source_url":"https://physionet.org/content/eicu-crd-demo/2.0.1/"},{"public_project_uuid":"5c7230ef-e9f2-4bb1-8f32-0e38a04fbaeb","slug":"cardioresp-response-orthostat","version":"1.0","core_doi":"10.13026/ehkk-8y70","version_doi":"10.13026/73w7-x881","is_latest_version":true,"short_description":"Dataset comprising the following physiological signals: blood pressure, respiratory rate, blood flow velocity in the middle cerebral arteries and tissue hemoglobin concentration in the prefrontal cortex.","publish_date":"2021-05-17","title":"Two-tiered response of cardiorespiratory-cerebrovascular networks to orthostatic challenge","abstract":"The authors used this dataset to evaluate the responsiveness of the cardiorespiratory-cerebrovascular networks by capturing linear and nonlinear interdependencies to postural changes. Ten young healthy adults participated in our study between December 2009 and February 2010 at Semmelweis University, Faculty of Medicine. Non-invasive measurements of arterial blood pressure (ABP), cardiac cycle duration (derived from ABP signals), breath-to-breath interval (capnography), cerebral blood flow velocity (BFV, recorded by transcranial Doppler sonography), and cerebral hemodynamics (hemoglobin signals monitored by near-infrared spectroscopy) were performed for 30-minutes in resting state, followed by a one-minute stand-up and a one-minute sit-down period. BFV and ABP signals were calibrated, and the dataset contains values in cm/s and mmHg, respectively. NIRS measurement yielded intensity values (raw signals) from which change of HbO (oxyhemoglobin), HbR (deoxyhemoglobin) and HbT (total hemoglobin) concentration in tissue can be calculated (in micromol/liter). Raw NIRS signals were synchronized to BFV, ABP and capnography time series yielding 33 minutes of resampled data (sampling frequency: ~ 3 Hz).
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":18635078,"compressed_storage_size":6970947,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/cardioresp-response-orthostat/1.0/"},{"public_project_uuid":"c738b4ea-0d04-482e-b9cb-1a36409d9329","slug":"mhd-effect-ecg-mri","version":"1.0.0","core_doi":"10.13026/7fxy-3653","version_doi":"10.13026/05td-jn37","is_latest_version":true,"short_description":"ECG signals were acquired in various MRI scanners to enable the study of the magnetohydrodynamic (MHD) effect. The MHD effect, which is caused by an interaction of the blood flow and the MRI’s high static magnetic field, superimposes the ECG signal.","publish_date":"2021-05-18","title":"Influence of the MHD effect on 12-lead and 3-lead ECGs recorded in 1T to 7T MRI scanners","abstract":"This ECG dataset was acquired in different magnetic resonance imaging (MRI) scanners to study the magnetohydrodynamic (MHD) effect. The MHD effect, which is caused an interaction of the MRI’s strong static magnetic field and the patient’s blood flow, superimposes the ECG signal during MRI exams. As a consequence, a detailed morphological analysis of the ECG (e.g. of the P wave, ST segment or the T wave) is prevented. The MHD effect might be a useful signal for extracting further physiological information, e.g. about blood flow or stroke volume. The 12-lead and 3-lead ECG signals were acquired from different subjects in various MRI scanners with magnetic field strengths ranging from 1T up to 7T. All signals were manually annotated.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":293413592,"compressed_storage_size":278146217,"resource_type":"Database","access_policy":"Open","topics":["cardiac mri","patient monitoring","magnetohydrodynamic","ecg","mri","mhd"],"source_url":"https://physionet.org/content/mhd-effect-ecg-mri/1.0.0/"},{"public_project_uuid":"a9188b3c-13b8-4bd2-bdb2-d043544c64af","slug":"nch-sleep","version":"0.1.0","core_doi":"10.13026/m463-tb74","version_doi":"10.13026/sq6b-9n68","is_latest_version":false,"short_description":"The NCH Sleep DataBank includes 3,984 pediatric sleep studies on 3,673 unique patients conducted at Nationwide Children's Hospital between 2017 and 2019. It contains polysomnography (PSG), clinical annotations, and longitudinal clinical data.","publish_date":"2021-05-19","title":"NCH Sleep DataBank: A Large Collection of Real-world Pediatric Sleep Studies with Longitudinal Clinical Data","abstract":"In order to accelerate research on pediatric sleep and its connection to health, Nationwide Children's Hospital (NCH) and Carnegie Mellon University (CMU) introduce the NCH Sleep DataBank. This dataset has 3,984 pediatric sleep studies on 3,673 unique patients conducted at NCH in Columbus, Ohio, USA between 2017 and 2019, along with the patient's longitudinal clinical data. The published Polysomnography (PSG) contains the patient's physiological signals as well as the technician's assessment of the sleep stages and descriptions of additional irregularities.
\n\nThe novelties of this dataset include: (1) Size: Its large size is suitable for discovering new scientific insights via data mining, (2) Patient population: It explicitly focuses on pediatric patients, (3) Clinical setting: The sleep studies were gathered in the real-world clinical setting at NCH as opposed to, for example, in a controlled clinical trial, and (4) Rich set of clinical data: The accompanying 5.6 million records of clinical data are extracted from the Electronic Health Record (EHR), and are separated into encounters, medications, measurements (e.g. body mass index), diagnoses, and procedures.
\n\nThe NCH Sleep DataBank is a valuable resource for advancing automatic sleep scoring and real-time sleep disorder prediction, among many other potential scientific discoveries. Accompanying code in Python to assist users in interacting with the dataset is published on GitHub.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":6675072816,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Open","topics":["ehr","ecg","eeg","polysomnography","sleep study","pediatrics","electronic health records","sleep disorders","clinical decision support"],"source_url":"https://physionet.org/content/nch-sleep/0.1.0/"},{"public_project_uuid":"31e05fb8-8474-486e-80a9-31ec91e461b3","slug":"heart-vector-origin-matlab","version":"1.0.0","core_doi":"10.13026/jsdm-1k72","version_doi":"10.13026/jfeg-4820","is_latest_version":true,"short_description":"The algorithm finds the heart vector origin point and constructs the time-coherent median beat. VCG origin point is defined as the electrically quiet or isoelectric state of the heart when the heart vector does not move in 3D space.","publish_date":"2021-05-25","title":"Heart Vector Origin Point Detection and Time-Coherent Median Beat Construction","abstract":"An appropriate selection of the heart vector origin point defines the ECG’s baseline, which then affects all subsequent ECG measurements. We introduce software capable of automatically identifying the heart vector origin point and constructing a time-coherent median beat. We define the heart vector electrocardiographic (VCG) origin point based on its electrophysiological and biophysical meaning: the electrically quiet, or isoelectric state of the heart when the heart vector does not move in 3D space. The electrically silent period is defined as the segment between the end of repolarization and the onset of P-wave (if present), or the beginning of the QRS complex.
\n\nThe definition of the VCG origin point determines the definition of the vector magnitude ECG baseline. Vector-Magnitude that is zeroed at the origin of the heart vector provides the most physiologically accurate cardiac beat. This is a fundamental step for any further global ECG and VCG measurements. Time-coherent median beat preserves the physiological sequence of cardiac activation and recovery. The time-coherent approach takes an easily detectable fiducial point (maximum |dV/dt|) as detected only on one ECG lead (e.g., X). Corresponding time-points on two other leads (e.g., Y and Z) are taken regardless of their morphological features. Such an approach significantly reduces the chance of error and preserves the physiological sequence of cardiac activation and recovery. The algorithm requires an input of a 10-second raw digital orthogonal ECG signal (XYZ leads).
","license":{"name":"MIT License"},"dua":null,"main_storage_size":4375277,"compressed_storage_size":3135451,"resource_type":"Software","access_policy":"Open","topics":["signal processing","baseline","vectorcardiogram","origin point","electrocardiogram","heart vector"],"source_url":"https://physionet.org/content/heart-vector-origin-matlab/1.0.0/"},{"public_project_uuid":"5ee23d02-757f-4943-914d-ff4ff0ee8a35","slug":"hd-semg","version":"1.0.0","core_doi":"10.13026/t9wv-d929","version_doi":"10.13026/ym7v-bh53","is_latest_version":false,"short_description":"We provide an open access dataset of High Density Surface Electromyogram (HD-sEMG). Our dataset can be used for both hand gesture classification-based neuroprosthetic control and EMG-force regression based proportional neuroprosthetic control.","publish_date":"2021-05-28","title":"Open Access Dataset and Toolbox of High-Density Surface Electromyogram Recordings","abstract":"We provide an open access dataset of High densitY Surface Electromyogram (HD-sEMG) Recordings (named "Hyser"), and a toolbox for neural interface research. We acquired data from 20 subjects with each subject participating in our experiment twice on separate days following the same experiment paradigm. Using our dataset, researchers can develop advanced techniques on pattern recognition of 34 hand gestures and regression between HD-sEMG and forces of five fingers. These techniques are essential for intuitive control of neuroprostheses and neuroexoskeletons. Our toolbox can be used to: (1) analyze each of the five datasets using standard benchmark methods and (2) decompose HD-sEMG signals into motor unit action potentials via independent component analysis.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":153294475081,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/hd-semg/1.0.0/"},{"public_project_uuid":"9e28eb58-ad35-4151-83bb-6e2c580531c5","slug":"mimic-iv-ed","version":"1.0","core_doi":"10.13026/1cjn-2370","version_doi":"10.13026/77z6-9w59","is_latest_version":false,"short_description":"A large database of emergency department admissions.","publish_date":"2021-06-03","title":"MIMIC-IV-ED","abstract":"MIMIC-ED is a large, freely available database of emergency department (ED) admissions at the Beth Israel Deaconess Medical Center between 2011 and 2019. As of MIMIC-ED v1.0, the database contains 448,972 ED stays. Vital signs, triage information, medication reconciliation, medication administration, and discharge diagnoses are available. All data are deidentified to comply with the Health Information Portability and Accountability Act (HIPAA) Safe Harbor provision. MIMIC-ED is intended to support a diverse range of education initiatives and research studies.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":142117642,"compressed_storage_size":142123437,"resource_type":"Database","access_policy":"Credentialed","topics":["ed","emergency","mimic","mimic-iv"],"source_url":"https://physionet.org/content/mimic-iv-ed/1.0/"},{"public_project_uuid":"d64777e4-c3cf-4036-afc4-7eb3970b26ff","slug":"radgraph","version":"1.0.0","core_doi":"10.13026/2f5n-pb92","version_doi":"10.13026/hm87-5p47","is_latest_version":true,"short_description":"RadGraph is a dataset of entities and relations in full-text chest X-ray radiology reports, which are obtained using a novel information extraction (IE) schema to capture clinically relevant information in a radiology report.","publish_date":"2021-06-03","title":"RadGraph: Extracting Clinical Entities and Relations from Radiology Reports","abstract":"RadGraph is a dataset of entities and relations in full-text radiology reports. We designed a novel information extraction (IE) schema to structure clinical information in a radiology report with four entities and three relations. Our train set consists of 500 MIMIC-CXR radiology reports annotated according to our schema by board-certified radiologists. Our test set consists of 50 MIMIC-CXR and 50 CheXpert reports, which are independently annotated by two board-certified radiologists. Additionally, we release annotations generated by a benchmark deep learning model that achieves a micro F1 of 0.82 (MIMIC-CXR test set) and 0.73 (CheXpert test set) on an evaluation metric for end-to-end relation extraction, where entity boundaries, entity types, and relation type must be correct. We use our model to automatically generate entity and relation labels across 220,763 MIMIC-CXR reports and 500 CheXpert reports, where annotations can be mapped to associated chest radiographs in the MIMIC-CXR and CheXpert datasets respectively. The dataset, which includes reports, entities, and relations, is de-identified according to the US Health Insurance Portability Act (HIPAA). This dataset is intended to support the development of natural language processing (NLP) methods for entity and relation extraction in radiology as well as enable multi-modal use cases that can leverage entities, relations, and associated radiographs.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1266726646,"compressed_storage_size":536490944,"resource_type":"Database","access_policy":"Credentialed","topics":["entity and relation extraction","natural language processing","graph","radiology","multi-modal"],"source_url":"https://physionet.org/content/radgraph/1.0.0/"},{"public_project_uuid":"6170ddcb-933e-4035-a0a7-974a39fefc19","slug":"q-pain","version":"1.0.0","core_doi":"10.13026/61xq-mj56","version_doi":"10.13026/2tdv-hj07","is_latest_version":true,"short_description":"Q-Pain, a medical QA dataset designed to enable the substitution of multiple different racial and gender \"profiles\" for patients and to evaluate whether bias is present when deciding whether to prescribe pain medication or not.","publish_date":"2021-06-11","title":"Q-Pain: A Question Answering Dataset to Measure Social Bias in Pain Management","abstract":"We introduce Q-Pain, a dataset for assessing bias in medical QA in the context of pain management. We developed 55 medical question-answer pairs across five different types of pain management: each question includes a detailed patient-specific medical scenario ("vignette") designed to enable the substitution of multiple different racial and gender "profiles" and to evaluate whether bias is present when answering whether or not to prescribe medication.
\n\nAlong with the dataset, we propose a new framework, including a sample experimental design, to measure the potential bias present during medical decision-making for patients with particular demographic profiles. We demonstrate its use in assessing two reference QA systems, GPT-2 and GPT-3, selected for their documented ability to answer questions given only a few examples. We hope that our dataset and framework will be used to assess bias across a wide variety of pain management QA systems.
","license":{"name":"Creative Commons Attribution-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":99899,"compressed_storage_size":29577,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/q-pain/1.0.0/"},{"public_project_uuid":"b9acb867-c8d4-4212-a6e4-daee38c91f32","slug":"ephnogram","version":"1.0.0","core_doi":"10.13026/kr2z-1881","version_doi":"10.13026/tjtq-5911","is_latest_version":true,"short_description":"An open-access database recorded during the EPHNOGRAM project, consisting of simultaneous electrocardiogram (ECG) and phonocardiogram (PCG) recordings from young healthy adults, during stress-test experiments.","publish_date":"2021-06-12","title":"EPHNOGRAM: A Simultaneous Electrocardiogram and Phonocardiogram Database","abstract":"The electro-phono-cardiogram (EPHNOGRAM) project focused on the development of low-cost and low-power devices for recording simultaneous electrocardiogram (ECG) and phonocardiogram (ECG) data, with auxiliary channels for capturing environmental audio noise, which could be used for PCG quality enhancement through signal processing. The current database, recorded by version 2.1 of the developed hardware, has been acquired from 24 healthy adults aged between 23 and 29 (average: 25.4 ± 1.9 years) in 30min stress-test sessions during resting, walking, running and biking conditions, using indoor fitness center equipment. The dataset also contains several 30s sample records acquired during rest conditions. This data is useful for simultaneous multi-modal analysis of ECG and PCG. It provides interesting insights into the inter-relationship between the mechanical and electrical mechanisms of the heart, under rest and physical activity.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":11281255939,"compressed_storage_size":6899377351,"resource_type":"Database","access_policy":"Open","topics":["stress-test","electrocardiogram","phonocardiogram"],"source_url":"https://physionet.org/content/ephnogram/1.0.0/"},{"public_project_uuid":"da53fb71-bd2b-4f7e-b86d-46fbc7aadcbe","slug":"covid-19-thermal","version":"1.0","core_doi":"10.13026/d9gh-4m08","version_doi":"10.13026/pgk4-gx55","is_latest_version":false,"short_description":"Thermal videos of people with positive and negative COVID-19 tests.","publish_date":"2021-06-21","title":"Upper body thermal images and associated clinical data from a pilot cohort study of COVID-19","abstract":"The prospective upper body thermal images SARS-CoV2 association study was designed to test the hypothesis that thermal videos may aid in the early diagnosis of COVID-19. The study recorded a set of measurements from 252 participants regarding PCR results, demographics, vital signs, participant activities, medications, respiratory symptoms, and a thermal video session where the volunteers performed simple breath-hold in four different positions. The acquired data may be used to test clinical association questions regarding temperature patterns, demographics, and vital signs. Furthermore, it could be valuable to develop new computer algorithms for extracting useful scientific information from thermal videos.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":1218284742,"compressed_storage_size":1084975024,"resource_type":"Database","access_policy":"Restricted","topics":["thermal videos","covid-19","sars-cov-2","clinical symptoms"],"source_url":"https://physionet.org/content/covid-19-thermal/1.0/"},{"public_project_uuid":"4321203e-0049-49c0-a638-5bf042579eb6","slug":"mimic-iii-clinical-action","version":"1.0.0","core_doi":"10.13026/6kfs-9v69","version_doi":"10.13026/kw00-z903","is_latest_version":true,"short_description":"Clinical action items annotated over MIMIC-III. 718 discharge summaries are labeled at a sentence- and character-level with multiple action labels including Appointment, Lab, Procedure, Medication, Imaging, Patient Instructions, and Other.","publish_date":"2021-06-21","title":"CLIP: A Dataset for Extracting Action Items for Physicians from Hospital Discharge Notes","abstract":"We created a dataset of clinical action items annotated over MIMIC-III. This dataset, which we call CLIP, is annotated by physicians and covers 718 discharge summaries, representing 107,494 sentences. Annotations were collected as character-level spans to discharge summaries after applying surrogate generation to fill in the anonymized templates from MIMIC-III text with faked data. We release these spans, their aggregation into sentence-level labels, and the sentence tokenizer used to aggregate the spans and label sentences. We also release the surrogate data generator, and the document IDs used for training, validation, and test splits, to enable reproduction. The spans are annotated with 0 or more labels of 7 different types, representing the different actions that may need to be taken: Appointment, Lab, Procedure, Medication, Imaging, Patient Instructions, and Other. We encourage the community to use this dataset to develop methods for automatically extracting clinical action items from discharge summaries.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":62212750,"compressed_storage_size":43110723,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimic-iii-clinical-action/1.0.0/"},{"public_project_uuid":"a8971424-d4e2-4ed7-8831-4c0ffe45f51c","slug":"cpsc2021","version":"1.0.0","core_doi":"10.13026/h37m-r024","version_doi":"10.13026/ksya-qw89","is_latest_version":true,"short_description":"CPSC2021 for paroxysmal atrial fibrillation events detection.","publish_date":"2021-06-21","title":"Paroxysmal Atrial Fibrillation Events Detection from Dynamic ECG Recordings: The 4th China Physiological Signal Challenge 2021","abstract":"Atrial fibrillation (AF) is the most frequent arrhythmia, but paroxysmal atrial fibrillation (PAF) often remains unrecognized. Early detection of PAF is of great value for AF surgery options, drug intervention and diagnosis and treatment of various clinical complications. Although accurate detection of paroxysmal AF is very important, there is currently no algorithm that can efficiently measure the onset and end of AF episode in dynamic or wearable ECGs. Previous AF detection algorithms usually focus on the classification of AF rhythm instead of locating the onsets and ends of AF episodes. Thus, the clinical significance for the personalized treatment and management of AF patients is limited. The identification of AF rhythm is also influenced by other abnormal rhythms in clinical applications.
\n\nThe 4th China Physiological Signal Challenge 2021 (CPSC 2021) aims to encourage the development of algorithms for searching the AF episodes in dynamic ECG records. A new dynamic ECG database was constructed to encourage the development of more efficient and robust algorithms for PAF detection. We also develop a new scoring metric to evaluate detection methods for PAF events.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":1392566649,"compressed_storage_size":1113858872,"resource_type":"Challenge","access_policy":"Open","topics":["event detection","paroxysmal atrial fibrillation"],"source_url":"https://physionet.org/content/cpsc2021/1.0.0/"},{"public_project_uuid":"aaa4d074-7c9b-487e-acc7-4d37b088dfc0","slug":"mimic-iv-demo-omop","version":"0.9","core_doi":"10.13026/2d25-8g07","version_doi":"10.13026/p1f5-7x35","is_latest_version":true,"short_description":"Preliminary work to transform a MIMIC-IV demo dataset to the OMOP Common Data Model","publish_date":"2021-06-22","title":"MIMIC-IV demo data in the OMOP Common Data Model","abstract":"In this project, the MIMIC-IV demo database was used to create an Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) instance. The transformation was built making use of template scripts and concept mappings from a previous project for converting MIMIC-III data to the OMOP CDM, while adjusting the logic and extending the mappings to be compliant with MIMIC-IV. The desired outcome was to establish and improve ways of transforming observational data collected in an intensive care environment to the OMOP CDM.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":76577583,"compressed_storage_size":10786504,"resource_type":"Database","access_policy":"Open","topics":["omop","common data model"],"source_url":"https://physionet.org/content/mimic-iv-demo-omop/0.9/"},{"public_project_uuid":"964e7710-d609-4f26-9a34-a67db5dcff4e","slug":"vindr-cxr","version":"1.0.0","core_doi":"10.13026/46y6-6x96","version_doi":"10.13026/3akn-b287","is_latest_version":true,"short_description":"VinDr-CXR: An open dataset of chest X-rays with radiologist's annotations","publish_date":"2021-06-22","title":"VinDr-CXR: An open dataset of chest X-rays with radiologist annotations","abstract":"We describe here a dataset of more than 100,000 chest X-ray scans that were retrospectively collected from two major hospitals in Vietnam. Out of this raw data, we release 18,000 images that were manually annotated by a total of 17 experienced radiologists with 22 local labels of rectangles surrounding abnormalities and 6 global labels of suspected diseases. The released dataset is divided into a training set of 15,000 and a test set of 3,000. Each scan in the training set was independently labeled by 3 radiologists, while each scan in the test set was labeled by the consensus of 5 radiologists. All images are in DICOM format and the labels from training and test sets are made publicly available.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":205970857003,"compressed_storage_size":151581123766,"resource_type":"Database","access_policy":"Credentialed","topics":["lesion detection","disease classification","computer vision","chest x-ray interpretation","deep learning"],"source_url":"https://physionet.org/content/vindr-cxr/1.0.0/"},{"public_project_uuid":"023d088a-5f1a-4671-b462-2d00cf7772bd","slug":"accelerometry-walk-climb-drive","version":"1.0.0","core_doi":"10.13026/fqq5-r521","version_doi":"10.13026/51h0-a262","is_latest_version":true,"short_description":"Labeled raw accelerometry data collected during outdoor walking, stair climbing, and driving for 32 healthy adults. Data were collected simultaneously at four body locations: left wrist, left hip, both ankles.","publish_date":"2021-06-26","title":"Labeled raw accelerometry data captured during walking, stair climbing and driving","abstract":"The database contains raw accelerometry data collected during outdoor walking, stair climbing, and driving for 32 healthy adults. Accelerometry data were collected simultaneously at four body locations: left wrist, left hip, left ankle, and right ankle, at a sampling frequency of 100 Hz. The 3-axial ActiGraph GT3X+ devices were used to collect the data. The data include labels of activity type performed (walking, descending stairs, ascending stairs, driving, clapping) for each time point of data collection. Basic demographic information of participants is also provided. All data are anonymized.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":797716553,"compressed_storage_size":184676079,"resource_type":"Database","access_policy":"Open","topics":["actigraph","accelerometers","walking activity","physical activity","activity monitor","accelerometer","driving activity","raw accelerometry data","accelerometry","activity meter","activity recognition"],"source_url":"https://physionet.org/content/accelerometry-walk-climb-drive/1.0.0/"},{"public_project_uuid":"4a500b62-49f3-4117-96b9-d8ac2eaf89ce","slug":"radnli-report-inference","version":"1.0.0","core_doi":"10.13026/qk1q-vq60","version_doi":"10.13026/mmab-c762","is_latest_version":true,"short_description":"A radiology NLI dataset introduced in the paper: Improving Factual Completeness and Consistency of Image-to-text Radiology Report Generation","publish_date":"2021-06-29","title":"RadNLI: A natural language inference dataset for the radiology domain","abstract":"The problem of natural language inference (NLI) determines whether a natural language hypothesis can be justifiably inferred from a natural language premise. NLI has attracted researchers to benchmark it in a number of settings including medical ones. While NLI datasets such as the MedNLI dataset exist for the clinical domain, systems trained with them do not generalize well to applications that require understanding of radiology reports. We therefore introduce an NLI dataset in the radiology domain, in which NLI information is annotated on sentences drawn from radiology reports. Sentence pairs in our dataset are sampled from MIMIC-CXR and we annotated them with NLI labels by two experts: one medical expert and one computer science expert. Each pair is annotated twice, swapping its premise and hypothesis, resulting in 960 pairs. The set is then split in half, resulting in 480 pair for a validation set and 480 pairs for a test set. We confirmed that a BERT-based NLI model trained with a distant supervision approach can achieve the accuracy of 77.8% on this test set.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":31919731,"compressed_storage_size":30878367,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/radnli-report-inference/1.0.0/"},{"public_project_uuid":"2e80102f-fe30-4f9a-b246-21703042db51","slug":"eda-rest-sedation","version":"1.0","core_doi":"10.13026/m060-cw82","version_doi":"10.13026/r9p1-bk90","is_latest_version":true,"short_description":"This database of pulse times and amplitudes from electrodermal activity was collected from 11 healthy volunteer subjects who were awake and at rest and 11 (different) healthy volunteer subjects who were under controlled propofol sedation.","publish_date":"2021-06-30","title":"Pulse Amplitudes from electrodermal activity collected from healthy volunteer subjects at rest and under controlled sedation","abstract":"This dataset of electrodermal activity was collected from 11 healthy volunteer subjects who were awake and at rest in seated position and 11 different healthy volunteers who were under controlled propofol sedation. For the awake and at rest subjects, the activity was recorded from each subject's non-dominant hand for one hour at 256 Hz. For the controlled propofol sedation subjects, the activity was recorded from each subject's left hand for about 3-4 hours at 500 Hz. From the raw data, EDA pulses were extracted and the pulse times and amplitudes reported. Electrodermal activity measures changing electrical conductance of the skin as an indicator of sweat gland activity. Sweat glands are a primitive part of the fight-or-flight response. These data were collected as part of a larger study to understand and build computational models for autonomic nervous system activity (including electrodermal activity) with approval from the Massachusetts Institute of Technology Committee on the Use of Humans as Experimental Subjects (COUHES) and the Massachusetts General Hospital Human Research Committee.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":331166,"compressed_storage_size":163507,"resource_type":"Database","access_policy":"Open","topics":["autonomic nervous system","anesthesia","electrodermal activity","sympathetic nervous system","sweat gland activity","propofol"],"source_url":"https://physionet.org/content/eda-rest-sedation/1.0/"},{"public_project_uuid":"7bdf1400-cc98-4af0-820e-5418b555ecbd","slug":"hmc-sleep-staging","version":"1.0.0","core_doi":"10.13026/t4w7-3k21","version_doi":"10.13026/gp48-ea60","is_latest_version":false,"short_description":"A collection of 154 whole-night PolySomnoGraphic (PSG) sleep recordings from the Haaglanden Medisch Centrum (HMC, The Netherlands) sleep center containing different traces of ExG activity and expert annotations regarding sleep staging task","publish_date":"2021-07-01","title":"Haaglanden Medisch Centrum sleep staging database","abstract":"A collection of 154 whole-night polysomnographic (PSG) sleep recordings (88 Male, 66 Female, mean Age of 53.8 ± 15.4) collected during 2018 at the Haaglanden Medisch Centrum (HMC, The Netherlands) sleep center. Patient recordings were randomly selected and include a heterogeneous population which was referred for PSG examination on the context of different sleep disorders. The dataset contains electroencephalographic (EEG), electrooculographic (EOG), chin electromyographic (EMG), and electrocardiographic (ECG) activity, as well as event annotations corresponding to scoring of sleep patterns (hypnogram) performed by sleep technicians at HMC. The dataset was collected as part of a study evaluating the generalization performance of an automatic sleep scoring algorithm across multiple heterogeneous datasets.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":17179343080,"compressed_storage_size":13876889182,"resource_type":"Database","access_policy":"Open","topics":["inter-database generalization","edf","deep learning","sleep staging"],"source_url":"https://physionet.org/content/hmc-sleep-staging/1.0.0/"},{"public_project_uuid":"9349124e-aeb3-4d28-b1d1-abf22a803cc5","slug":"chest-imagenome","version":"1.0.0","core_doi":"10.13026/9ths-ht80","version_doi":"10.13026/wv01-y230","is_latest_version":true,"short_description":"The Chest ImaGenome dataset is a scene graph dataset with additional chronological comparison relations for chest X-rays. It is automatically derived from the MIMIC-CXR dataset. A manually annotated gold standard is also available for 500 patients.","publish_date":"2021-07-13","title":"Chest ImaGenome Dataset","abstract":"In recent years, with the release of multiple large datasets, automatic interpretation of chest X-ray (CXR) images with deep learning models have become feasible for specific abnormalities or for generating preliminary reports. However, despite reports of performance reaching similar levels to that of radiologists, a quantitative evaluation of the explainability of these models is hampered by the lack of locally labeled datasets for different findings. With the exception of a few human-labeled small-scale datasets for specific findings, such as pneumonia and pneumothorax, most of the CXR deep learning models to date are trained on global "weak" labels extracted from text reports, or trained via a joint image and unstructured text learning strategy. In our work, a joint rule-based natural language processing (NLP) and CXR atlas-based bounding box detection pipeline are used to automatically label 242072 frontal MIMIC CXRs locally. Inspired by the Visual Genome effort in the computer vision community [20], we constructed the first Chest ImaGenome dataset with a scene graph data structure to describe the data. Through a radiologist constructed CXR ontology, the annotations for each CXR are connected as an anatomy-centered scene graph, useful for image-level reasoning and multimodal fusion applications. Overall, our dataset contributes significantly to the research community by providing 1) 1,256 combinations of relation annotations between 29 CXR anatomical locations (objects with bounding box coordinates) and their attributes, structured as a scene graph per image, 2) over 670,000 localized comparison relations (for improved, worsened, or no change) between the anatomical locations across sequential exams, as well as 3) a manually annotated gold standard scene graph dataset from 500 unique patients.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":6436570666,"compressed_storage_size":1553519249,"resource_type":"Database","access_policy":"Credentialed","topics":["scene graph","visual question answering","visual dialogue","machine learning","object detection","deep learning","disease progression","multimodal","semantic reasoning","bounding box","radiology","relation extraction","chest x-ray","knowledge graph","cxr","chest","explainability","reasoning"],"source_url":"https://physionet.org/content/chest-imagenome/1.0.0/"},{"public_project_uuid":"30d73d2e-55ee-482f-83bd-dc967ff25d70","slug":"cpap-data-canterbury","version":"1.0.0","core_doi":"10.13026/33rx-gh20","version_doi":"10.13026/ckz9-ph15","is_latest_version":false,"short_description":"A pressure and flow dataset was collected from a trial of 30 adults at the University of Canterbury undergoing CPAP therapy for a variety of instructed breath rates at PEEP levels of 4cmH2O and 7cmH2O.","publish_date":"2021-07-19","title":"CPAP Pressure and Flow Data from a Local Trial of 30 Adults at the University of Canterbury","abstract":"A pressure and flow dataset for CPAP (continuous positive airway pressure) breathing obtained from 30 subjects for model-based identification of patient-specific lung mechanics using a specially designed sensor system comprising an array of differential pressure sensors (Sensirion AG SDP816-125PA) and gauge pressure sensors (NXP MPVZ4006GW7U). Relevant medical information was collected using a questionnaire, including: sex; age; weight; height; smoking history; and history of asthma. Subjects were tasked with breathing at five different rates (including passive) from very slow to very fast breathing, matched to an online pacing sound and video, at two different levels of CPAP (4 and 7 cmH2O) for between 50 and 180 seconds depending on rate. Each data set for a breathing rate comprises ~17 breaths, including rest periods between breathing rates and CPAP levels. Data with significant observed sensor error was removed during processing. A better-fitted mask would be expected to reduce error in fully capturing expiration by reducing leaks around the masks seal with the face. The sensors measure inspiration and expiration using separate unidirectional differential flow sensors, which are combined in post-processing to for complete breathing data.
","license":{"name":"Creative Commons Attribution-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":244438237,"compressed_storage_size":93230631,"resource_type":"Database","access_policy":"Open","topics":["cpap","respiratory mechanics","biomedical engineering","respiratory modelling","peep","pulmonary mechanics"],"source_url":"https://physionet.org/content/cpap-data-canterbury/1.0.0/"},{"public_project_uuid":"f10b0cb0-8ac7-404d-9cf1-b69783a6793f","slug":"autonomic-aging-cardiovascular","version":"1.0.0","core_doi":"10.13026/mwad-va85","version_doi":"10.13026/2hsy-t491","is_latest_version":true,"short_description":"This database contains resting recordings of ECG and continuous noninvasive blood pressure of 1,104 healthy volunteers","publish_date":"2021-07-30","title":"Autonomic Aging: A dataset to quantify changes of cardiovascular autonomic function during healthy aging","abstract":"Autonomic function regulating blood pressure and cardiac rhythm progressively declines with increasing age. Impaired cardiovascular control promotes a variety of age-related conditions such as dementia, or Alzheimer’s disease. This study aims to provide a database of high-resolution biological signals to describe the effect of healthy aging on cardiovascular regulation. Electrocardiogram and continuous non-invasive blood pressure signals were recorded simultaneously at rest in 1,121 healthy volunteers.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":6288796748,"compressed_storage_size":5578641955,"resource_type":"Database","access_policy":"Open","topics":["blood pressure","age","cardiovascular","baroreflex","heart rate variability"],"source_url":"https://physionet.org/content/autonomic-aging-cardiovascular/1.0.0/"},{"public_project_uuid":"745c9677-67f3-41b7-8682-fdb8e8a34eaa","slug":"propofol-anesthesia-dynamics","version":"1.0","core_doi":"10.13026/gmav-2367","version_doi":"10.13026/2rbc-1r03","is_latest_version":true,"short_description":"Multimodal point process indices for heart rate variability and electrodermal activity for 9 subjects who are undergoing a controlled propofol sedation experiment where the concentration was increased and then decreased in stages.","publish_date":"2021-07-30","title":"Behavioral and autonomic dynamics during propofol-induced unconsciousness","abstract":"During general anesthesia, both behavioral and autonomic changes are caused by the administration of anesthetics such as propofol. Propofol produces unconsciousness by creating highly structured oscillations in brain circuits. The anesthetic also has autonomic effects due to its actions as a vasodilator and myocardial depressant. Understanding how autonomic dynamics change in relation to propofol-induced unconsciousness is an important scientific and clinical question since anesthesiologists often infer changes in level of unconsciousness from changes in autonomic dynamics.
\n\nTherefore, we present a dataset combining physiology-based indices for heart rate variability and electrodermal activity with a robust statistical tool to compare behavioral and multimodal autonomic changes before, during, and after propofol-induced unconsciousness. This dataset was recorded from nine healthy volunteers during computer-controlled administration of propofol.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":7092005335,"compressed_storage_size":1415612550,"resource_type":"Database","access_policy":"Open","topics":["anesthesia","autonomic nervous system","propofol","heart rate variability","electrodermal activity"],"source_url":"https://physionet.org/content/propofol-anesthesia-dynamics/1.0/"},{"public_project_uuid":"37e3b950-f475-48f8-9f7a-5e8c088723bb","slug":"covid-19-thermal","version":"1.1","core_doi":"10.13026/d9gh-4m08","version_doi":"10.13026/wfr2-5973","is_latest_version":true,"short_description":"Thermal videos of people with positive and negative COVID-19 tests.","publish_date":"2021-08-16","title":"Upper body thermal images and associated clinical data from a pilot cohort study of COVID-19","abstract":"The prospective upper body thermal images SARS-CoV2 association study was designed to test the hypothesis that thermal videos may aid in the early diagnosis of COVID-19. The study recorded a set of measurements from 252 participants regarding PCR results, demographics, vital signs, participant activities, medications, respiratory symptoms, and a thermal video session where the volunteers performed simple breath-hold in four different positions. The acquired data may be used to test clinical association questions regarding temperature patterns, demographics, and vital signs. Furthermore, it could be valuable to develop new computer algorithms for extracting useful scientific information from thermal videos.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":1218369731,"compressed_storage_size":1085049550,"resource_type":"Database","access_policy":"Restricted","topics":["covid-19","thermal videos","sars-cov-2","clinical symptoms"],"source_url":"https://physionet.org/content/covid-19-thermal/1.1/"},{"public_project_uuid":"50d8cf27-d735-4ab3-98ea-5b32a9c8995a","slug":"vindr-spinexr","version":"1.0.0","core_doi":"10.13026/mjat-sv46","version_doi":"10.13026/q45h-5h59","is_latest_version":true,"short_description":"VinDr-SpineXR: A large annotated medical image dataset for spinal lesions detection and classification from radiographs","publish_date":"2021-08-24","title":"VinDr-SpineXR: A large annotated medical image dataset for spinal lesions detection and classification from radiographs","abstract":"Radiographs are used as the most critical imaging tool for identifying spine anomalies in clinical practice [1]. The evaluation of spinal bone lesions, however, is a challenging task for radiologists. To the best of our knowledge, no existing studies are devoted to developing and evaluating a comprehensive system for classifying and localizing multiple spine lesions from X-ray scans. The lack of large-scale spine X-ray datasets with high-quality images and human expert annotations is the key obstacle. To fill this gap, we introduce a large-scale annotated medical image dataset for spinal lesion detection and classification from radiographs. The dataset, called VinDr-SpineXR, contains 10,466 spine X-ray images from 5,000 studies, each of which is manually annotated with 13 types of abnormalities by an experienced radiologist with bounding boxes around abnormal findings. This is the largest dataset to date that provides radiologist's bounding-box annotations for developing supervised-learning algorithms for spine X-ray analysis.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":38632031432,"compressed_storage_size":35797245250,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/vindr-spinexr/1.0.0/"},{"public_project_uuid":"672c08ac-eb7d-45cb-b83a-9ba686b0fd3e","slug":"electrodermal-activity","version":"2.0","core_doi":"10.13026/qgx3-3b28","version_doi":"10.13026/136t-7g98","is_latest_version":true,"short_description":"This database of electrodermal activity was collected from 11 healthy volunteer subjects who were awake and at rest in seated position and 11 (different) healthy volunteer subjects who were under computer-controlled propofol sedation.","publish_date":"2021-08-26","title":"Electrodermal Activity of Healthy Volunteers while Awake and at Rest","abstract":"This dataset of electrodermal activity was collected from 11 healthy volunteer subjects who were awake and at rest in seated position. The activity was recorded from each subject's non-dominant hand for one hour at 256 Hz. Electrodermal activity measures changing electrical conductance of the skin as an indicator of sweat gland activity. Sweat glands are a primitive part of the fight-or-flight response. These data were collected as part of a larger study to understand and build computational models for autonomic nervous system activity (including electrodermal activity) with approval from the Massachusetts Institute of Technology Committee on the Use of Humans as Experimental Subjects (COUHES). The second dataset of electrodermal activity was collected from 11 healthy volunteer subjects who were under computer-controlled propofol sedation. The activity was recorded from each subject's left hand for about 3 hours at 500 Hz. These data were collected with approval from the Massachusetts General Hospital Human Research Committee.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1588592776,"compressed_storage_size":488491234,"resource_type":"Database","access_policy":"Credentialed","topics":["sweat gland activity","autonomic nervous system","anesthesia","sympathetic nervous system","electrodermal activity","propofol"],"source_url":"https://physionet.org/content/electrodermal-activity/2.0/"},{"public_project_uuid":"56822999-ac09-488b-9dd7-f84c18e45379","slug":"nch-sleep","version":"1.0.0","core_doi":"10.13026/m463-tb74","version_doi":"10.13026/fp4y-6m41","is_latest_version":false,"short_description":"The NCH Sleep DataBank includes 3,984 pediatric sleep studies on 3,673 unique patients conducted at Nationwide Children's Hospital between 2017 and 2019. It contains polysomnography (PSG), clinical annotations, and longitudinal clinical data.","publish_date":"2021-08-28","title":"NCH Sleep DataBank: A Large Collection of Real-world Pediatric Sleep Studies with Longitudinal Clinical Data","abstract":"In order to accelerate research on pediatric sleep and its connection to health, Nationwide Children's Hospital (NCH) and Carnegie Mellon University (CMU) introduce the NCH Sleep DataBank. This dataset has 3,984 pediatric sleep studies on 3,673 unique patients conducted at NCH in Columbus, Ohio, USA between 2017 and 2019, along with the patient's longitudinal clinical data. The published Polysomnography (PSG) contains the patient's physiological signals as well as the technician's assessment of the sleep stages and descriptions of additional irregularities.
\n\nThe novelties of this dataset include: (1) Size: Its large size is suitable for discovering new scientific insights via data mining, (2) Patient population: It explicitly focuses on pediatric patients, (3) Clinical setting: The sleep studies were gathered in the real-world clinical setting at NCH as opposed to, for example, in a controlled clinical trial, and (4) Rich set of clinical data: The accompanying 5.6 million records of clinical data are extracted from the Electronic Health Record (EHR), and are separated into encounters, medications, measurements (e.g. body mass index), diagnoses, and procedures.
\n\nThe NCH Sleep DataBank is a valuable resource for advancing automatic sleep scoring and real-time sleep disorder prediction, among many other potential scientific discoveries. Accompanying code in Python to assist users in interacting with the dataset is published on GitHub.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":0,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Open","topics":["ehr","eeg","ecg","clinical decision support","pediatrics","sleep disorders","sleep study","electronic health records","polysomnography"],"source_url":"https://physionet.org/content/nch-sleep/1.0.0/"},{"public_project_uuid":"3bff3487-6fe0-4c1b-9ed4-d30bac6e3cf2","slug":"12-lead-ecg-viewer-matlab","version":"1.0.0","core_doi":"10.13026/64cv-ah78","version_doi":"10.13026/sm8m-v308","is_latest_version":true,"short_description":"Clinical Viewer of raw digital 12-lead ECG file (ECG file in .txt format).","publish_date":"2021-08-30","title":"Lightweight 12-lead ECG viewer for MATLAB","abstract":"Electrocardiograms (ECGs) provide time-varying measurements of the electrical activity of the heart. These signals can provide insights into underlying aspects of heart health and disease. Here we provide software for viewing 12-lead ECG files (in .txt format) in a form familiar to any clinician or ECG expert, facilitating manual measurements of intervals and amplitudes. An executable version of the software, without the need for Matlab installation, can be downloaded separately. The input file is a 10-second 12 lead ECG text file, the output from the Magellan ECG Research Workstation Software Version 2.0 (GE Marquette Medical Systems, USA).
","license":{"name":"MIT License"},"dua":null,"main_storage_size":418937,"compressed_storage_size":246771,"resource_type":"Software","access_policy":"Open","topics":["routine clinical ecg","electrocardiogram","viewer","clinical 12-lead ecg"],"source_url":"https://physionet.org/content/12-lead-ecg-viewer-matlab/1.0.0/"},{"public_project_uuid":"ba4426ff-127e-4594-8746-431999779129","slug":"rr-interval-healthy-subjects","version":"1.0.0","core_doi":"10.13026/c61a-z467","version_doi":"10.13026/51yd-d219","is_latest_version":true,"short_description":"This database contains RR intervals time series from healthy subjects aged between 1month and 55 years","publish_date":"2021-08-30","title":"RR interval time series from healthy subjects","abstract":"Heart rate variability (HRV) is mainly controlled by the autonomic nervous system (ANS) interacting with receptors on the sinoatrial node (SAN). HRV is influenced by many several factors such as chemical, hormonal and neural modulations, circadian changes, exercise, emotions, posture, and preload. The heart rate adaptation to changing factors is carried out by the activity of a variety of regulatory subsystems and results in a complex linear and non-linear time behaviour, which changes with age, gender and pathological conditions. Many studies explore an age range of 1-99 years. This study focused on the development of a database of long-term time series of healthy volunteers aged between 1 month and 55 years.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":92213009,"compressed_storage_size":17709181,"resource_type":"Database","access_policy":"Open","topics":["rr time series","healthy subjects","heart rate variability","from newly born to adult"],"source_url":"https://physionet.org/content/rr-interval-healthy-subjects/1.0.0/"},{"public_project_uuid":"6ff46239-039c-4994-bbb9-808001cd3369","slug":"ecg-ppg-simulator-arrhythmia","version":"1.3.0","core_doi":"10.13026/xf56-pm38","version_doi":"10.13026/c5qq-2g72","is_latest_version":false,"short_description":"A model is capable of simulating sinus rhythm, atrial fibrillation and ectopic beats in ECGs and PPGs as well as extreme bradycardia and ventricular tachycardia in PPGs. Different types of noises and artifacts can also be added to the waveforms.","publish_date":"2021-08-31","title":"Model for Simulating ECG and PPG Signals with Arrhythmia Episodes","abstract":"The simulation of electrocardiographic (ECG) and photoplethysmographic (PPG) signals with cardiac arrhythmias is important when developing and testing arrhythmia detectors. This software is capable of simulating sinus rhythm, episodes of atrial fibrillation and atrial premature beats in ECGs and PPGs as well as extreme bradycardia and ventricular tachycardia in PPGs. Different types of noise and artifacts can be added to simulated signals to improve realism and mimic the signals acquired in free-living conditions. Two versions are provided: a binary version for Windows 10 and Matlab source code.
","license":{"name":"GNU General Public License version 3"},"dua":null,"main_storage_size":5137665384,"compressed_storage_size":4938661597,"resource_type":"Software","access_policy":"Open","topics":["atrial fibrillation","motion artifacts","noise","arrhythmia","bradycardia","tachycardia","simulation","ecg","ppg","detection"],"source_url":"https://physionet.org/content/ecg-ppg-simulator-arrhythmia/1.3.0/"},{"public_project_uuid":"01bb63cf-d8bd-4d88-979b-ed13aecb0a5e","slug":"icu-infection-zigong-fourth","version":"1.0","core_doi":"10.13026/cgzc-vp58","version_doi":"10.13026/gz5h-e561","is_latest_version":false,"short_description":"Routinely collected data from critical care units at Zigong Fourth People’s Hospital, Sichuan, China for patients admitted between January 2019 and December 2020","publish_date":"2021-09-09","title":"Critical care database comprising patients with infection at Zigong Fourth People's Hospital","abstract":"Patients treated in the intensive care unit (ICU) are closely monitored and receive intensive treatment. Such aggressive monitoring and treatment will generate high-granularity data from both electronic healthcare records and nursing chart. These data not only provide infrastructure for daily clinical practice, but also can help to inform clinical researches. It is technically challenging to integrate and cleanse medical data from a variety of sources. Although there are several open access critical care databases from western countries, there is a lack of this kind of database for Chinese adult patients. We established a critical care database involving patients with infection. A large proportion of these patients have sepsis and/or septic shock. High-granularity data comprising laboratory findings, baseline characteristics, medications, ICD code, nursing charts and follow-up results were integrated to generate a comprehensive database. The database can be utilized for a variety of clinical studies.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":491300142,"compressed_storage_size":94288071,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/icu-infection-zigong-fourth/1.0/"},{"public_project_uuid":"10c31d11-1309-425a-81b6-e3c3e988ef6b","slug":"entity-bert","version":"1.0.0","core_doi":"10.13026/vagp-v123","version_doi":"10.13026/kwdm-6a56","is_latest_version":false,"short_description":"Pretraining of models with a broad representation of biomedical terminology (PubMedBERT) on MIMIC-III corpus along with or without a novel entity-centric masking strategy.","publish_date":"2021-09-09","title":"EntityBERT: BERT-based Models Pretrained on MIMIC-III with or without Entity-centric Masking Strategy for the Clinical Domain","abstract":"Transformer-based neural language models have led to breakthroughs for a variety of natural language processing (NLP) tasks. However, most models are pretrained on general domain data. We propose a methodology to produce a model focused on the clinical domain: continued pretraining of a model with a broad representation of biomedical terminology (PubMedBERT) on a clinical corpus (MIMIC-III) along with a novel entity-centric masking strategy to infuse domain knowledge in the learning process.
\n\nWe curated the MIMIC-III corpus by annotating events (including diseases/disorders, signs/symptoms, medications, anatomical sites, and procedures) and time expressions (e.g. "yesterday", "this weekend", "02/31/2028"(an example date)) with special markers. Marked events and time expressions are randomly chosen together with other words in a certain ratio to be masked for training the entity-centric mask language model. Therefore, the models are infused with clinical entity information and good for entity-related clinical NLP tasks.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1322265301,"compressed_storage_size":1225860639,"resource_type":"Model","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/entity-bert/1.0.0/"},{"public_project_uuid":"3bf575e8-4e2d-40b1-aa64-cd7cd8bd806a","slug":"nch-sleep","version":"2.0.0","core_doi":"10.13026/m463-tb74","version_doi":"10.13026/pnkb-9r03","is_latest_version":false,"short_description":"The NCH Sleep DataBank includes 3,984 pediatric sleep studies on 3,673 unique patients conducted at Nationwide Children's Hospital between 2017 and 2019. It contains polysomnography (PSG), clinical annotations, and longitudinal clinical data.","publish_date":"2021-09-09","title":"NCH Sleep DataBank: A Large Collection of Real-world Pediatric Sleep Studies with Longitudinal Clinical Data","abstract":"In order to accelerate research on pediatric sleep and its connection to health, Nationwide Children's Hospital (NCH) and Carnegie Mellon University (CMU) introduce the NCH Sleep DataBank. This dataset has 3,984 pediatric sleep studies on 3,673 unique patients conducted at NCH in Columbus, Ohio, USA between 2017 and 2019, along with the patient's longitudinal clinical data. The published Polysomnography (PSG) contains the patient's physiological signals as well as the technician's assessment of the sleep stages and descriptions of additional irregularities.
\n\nThe novelties of this dataset include: (1) Size: Its large size is suitable for discovering new scientific insights via data mining, (2) Patient population: It explicitly focuses on pediatric patients, (3) Clinical setting: The sleep studies were gathered in the real-world clinical setting at NCH as opposed to, for example, in a controlled clinical trial, and (4) Rich set of clinical data: The accompanying 5.6 million records of clinical data are extracted from the Electronic Health Record (EHR), and are separated into encounters, medications, measurements (e.g. body mass index), diagnoses, and procedures.
\n\nThe NCH Sleep DataBank is a valuable resource for advancing automatic sleep scoring and real-time sleep disorder prediction, among many other potential scientific discoveries. Accompanying code in Python to assist users in interacting with the dataset is published on GitHub.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":0,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Restricted","topics":["polysomnography","eeg","sleep study","ecg","ehr","pediatrics","sleep disorders","electronic health records","clinical decision support"],"source_url":"https://physionet.org/content/nch-sleep/2.0.0/"},{"public_project_uuid":"9993e0a6-b111-4356-90d6-b6aec212f13c","slug":"nch-sleep","version":"3.0.0","core_doi":"10.13026/m463-tb74","version_doi":"10.13026/ema4-bv32","is_latest_version":false,"short_description":"The NCH Sleep DataBank includes 3,984 pediatric sleep studies on 3,673 unique patients conducted at Nationwide Children's Hospital between 2017 and 2019. It contains polysomnography (PSG), clinical annotations, and longitudinal clinical data.","publish_date":"2021-09-13","title":"NCH Sleep DataBank: A Large Collection of Real-world Pediatric Sleep Studies with Longitudinal Clinical Data","abstract":"In order to accelerate research on pediatric sleep and its connection to health, Nationwide Children's Hospital (NCH) and Carnegie Mellon University (CMU) introduce the NCH Sleep DataBank. This dataset has 3,984 pediatric sleep studies on 3,673 unique patients conducted at NCH in Columbus, Ohio, USA between 2017 and 2019, along with the patient's longitudinal clinical data. The published Polysomnography (PSG) contains the patient's physiological signals as well as the technician's assessment of the sleep stages and descriptions of additional irregularities.
\n\nThe novelties of this dataset include: (1) Size: Its large size is suitable for discovering new scientific insights via data mining, (2) Patient population: It explicitly focuses on pediatric patients, (3) Clinical setting: The sleep studies were gathered in the real-world clinical setting at NCH as opposed to, for example, in a controlled clinical trial, and (4) Rich set of clinical data: The accompanying 5.6 million records of clinical data are extracted from the Electronic Health Record (EHR), and are separated into encounters, medications, measurements (e.g. body mass index), diagnoses, and procedures.
\n\nThe NCH Sleep DataBank is a valuable resource for advancing automatic sleep scoring and real-time sleep disorder prediction, among many other potential scientific discoveries. Accompanying code in Python to assist users in interacting with the dataset is published on GitHub.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":2277743743533,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":["electronic health records","polysomnography","ecg","sleep disorders","pediatrics","sleep study","eeg","clinical decision support","ehr"],"source_url":"https://physionet.org/content/nch-sleep/3.0.0/"},{"public_project_uuid":"c4b3e7cf-cbf0-4c51-900b-32af205c5862","slug":"calcium-imaging-sleep-state","version":"1.0.0","core_doi":"10.13026/q34j-yz38","version_doi":"10.13026/jzqa-j937","is_latest_version":false,"short_description":"Wide-field calcium imaging database that consists of annotated sleep recording collected from transgenic mice at Washington University of St Louis School of Medicine.","publish_date":"2021-09-20","title":"Wide-field calcium imaging sleep state database","abstract":"A collection of wide-field calcium imaging (WFCI) sleep and wake recordings collected from transgenic mice expressing GCaMP6f in excitatory neurons. Each mouse underwent a three-hour undisturbed WFCI recording session where wake, NREM and REM sleep was recorded. Each WFCI recording is manually scored by sleep scoring experts in 10-second epochs as wake, NREM or REM by use of adjunct EEG/EMG. The dataset contains annotated WFCI recordings, brain mask and the Paxinos atlas used for defining the brain regions. The dataset was collected as part of a study evaluating a deep learning-based automated sleep state classification method.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":225878017293,"compressed_storage_size":225516582855,"resource_type":"Database","access_policy":"Open","topics":["machine learning","wide-field calcium imaging","sleep","sleep state classification","sleep staging"],"source_url":"https://physionet.org/content/calcium-imaging-sleep-state/1.0.0/"},{"public_project_uuid":"1e1e04c6-3859-44fb-8d0d-4019c8099020","slug":"ffa-ir-medical-report","version":"1.0.0","core_doi":"10.13026/k5rp-9h43","version_doi":"10.13026/ccbh-z832","is_latest_version":false,"short_description":"Benchmark dataset for report generation based on fundus fluorescein angiography images and reports.","publish_date":"2021-09-22","title":"FFA-IR: Towards an Explainable and Reliable Medical Report Generation Benchmark","abstract":"Automatic medical report generation (MRG) towards describing life-threatening lesions from given medical images, such as Chest X-ray and Fundus Fluorescein Angiography (FFA), has been a long-standing research topic in machine learning and automatic medical diagnosis fields. However, existing MRG benchmarks only provide medical images and free-text reports without explainable annotations and reliable evaluation tools, hindering the current research advances from two aspects: First, existing methods can only predict reports without accurate explanation, undermining the trustworthiness of the diagnostic methods; Second, the comparison between predicted reports from MRG methods is unreliable based on the natural language generation (NLG) metrics.
\n\nTo address these issues, we propose an explainable and reliable MRG benchmark based on FFA Images and Reports (FFA-IR). Specifically, our FFA-IR dataset is featured from the following aspects: 1) Large-scale medical dataset. FFA-IR collects 10,790 reports along with 1,048,584 FFA images from clinical practice. 2) Explainable annotation. FFA-IR annotates 46 categories of lesions with a total of 12,166 regions. 3) Bilingual reports. FFA-IR provides both English and Chinese reports for each case. We hope that our FFA-IR can significantly advance research from both vision-and-language and medicine fields and improve the conventional retinal disease diagnosis procedure.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":78867754983,"compressed_storage_size":77675381390,"resource_type":"Database","access_policy":"Credentialed","topics":["fundus fluorescein angiography","explainable and reliable evaluation","vision and language","medical report generation"],"source_url":"https://physionet.org/content/ffa-ir-medical-report/1.0.0/"},{"public_project_uuid":"d9054954-87a5-4b5c-8868-7f0de7cd7b8f","slug":"hmc-sleep-staging","version":"1.0.1","core_doi":"10.13026/t4w7-3k21","version_doi":"10.13026/7egw-0p30","is_latest_version":false,"short_description":"A collection of 154 whole-night PolySomnoGraphic (PSG) sleep recordings from the Haaglanden Medisch Centrum (HMC, The Netherlands) sleep center containing different traces of ExG activity and expert's scorings of sleep stages","publish_date":"2021-09-27","title":"Haaglanden Medisch Centrum sleep staging database","abstract":"A collection of 154 whole-night polysomnographic (PSG) sleep recordings (88 Male, 66 Female, mean Age of 53.8 ± 15.4) collected during 2018 at the Haaglanden Medisch Centrum (HMC, The Netherlands) sleep center. Patient recordings were randomly selected and include a heterogeneous population which was referred for PSG examination on the context of different sleep disorders. The dataset contains electroencephalographic (EEG), electrooculographic (EOG), chin electromyographic (EMG), and electrocardiographic (ECG) activity, as well as event annotations corresponding to scoring of sleep patterns (hypnogram) performed by sleep technicians at HMC. The dataset was collected as part of a study evaluating the generalization performance of an automatic sleep scoring algorithm across multiple heterogeneous datasets.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":17179343080,"compressed_storage_size":13876889182,"resource_type":"Database","access_policy":"Open","topics":["sleep staging","psg","inter-database generalization","edf","deep learning"],"source_url":"https://physionet.org/content/hmc-sleep-staging/1.0.1/"},{"public_project_uuid":"5af2c68d-3647-4c8a-a05d-10dc82d27b7b","slug":"reflacx-xray-localization","version":"1.0.0","core_doi":"10.13026/qhft-e228","version_doi":"10.13026/e0dj-8498","is_latest_version":true,"short_description":"This dataset contains 3032 cases of eye-tracking data collected while five radiologists dictated reports for frontal chest x-rays, synchronized timestamped dictation transcription, and manual labels for validation of localization of abnormalities.","publish_date":"2021-09-27","title":"REFLACX: Reports and eye-tracking data for localization of abnormalities in chest x-rays","abstract":"Labels localizing anomalies are rare in current chest x-rays datasets. We collected a dataset, using a method that can potentially be scaled up, to be a proof-of-concept for collecting implicit localization data through an eye-tracker. This dataset, named REFLACX, provides eye-tracking data collected while radiologists dictated reports for frontal chest x-rays from the MIMIC-CXR dataset, paired with the timestamped transcription of the dictation. The dataset contains 3,032 cases of synchronized eye-tracking and transcription pairs labeled by five radiologists. For each case, we also provide other labels for validating algorithms derived from this dataset. These labels include image-level labels of anomalies found during the reading, ellipses localizing these anomalies, and bounding boxes around lungs and heart. A subset of 109 chest x-rays has readings from all radiologists to estimate the variability of the provided labels.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":8831431386,"compressed_storage_size":631037124,"resource_type":"Database","access_policy":"Restricted","topics":["radiology report","computer vision","reflacx","fixations","chest x-rays","radiology","gaze","machine learning","eye tracking","deep learning"],"source_url":"https://physionet.org/content/reflacx-xray-localization/1.0.0/"},{"public_project_uuid":"af080e89-db59-4f18-9199-fb527599c927","slug":"wfdb-python","version":"3.4.1","core_doi":"10.13026/jt5e-gq15","version_doi":"10.13026/egpf-2788","is_latest_version":false,"short_description":"Tools for working with waveforms in Python.","publish_date":"2021-10-04","title":"Waveform Database Software Package (WFDB) for Python","abstract":"Physiological waveforms - such as electrocardiograms (ECG), electroencephalograms (EEG), electromyograms (EMG) - are generated during the course of routine care. These signals contain information that can be used to understand underlying conditions of health. Effective processing and analysis of physiological data requires specialized software. The waveform-database (WFDB) package for Python is a library of tools for reading, writing, and processing physiological signals and annotations, written in the Python programming language. Core components of the Python package are loosely based on the specifications of the original C-language WFDB software. We aim to implement as many of the core WFDB features as possible, with user-friendly APIs. Additional features are added over time.
","license":{"name":"MIT License"},"dua":null,"main_storage_size":244856323,"compressed_storage_size":103941864,"resource_type":"Software","access_policy":"Open","topics":["wfdb","waveform","python"],"source_url":"https://physionet.org/content/wfdb-python/3.4.1/"},{"public_project_uuid":"31f35da5-c8b9-4938-8d06-deba4bdef6e8","slug":"nch-sleep","version":"3.1.0","core_doi":"10.13026/m463-tb74","version_doi":"10.13026/p2rp-sg37","is_latest_version":true,"short_description":"The NCH Sleep DataBank includes 3,984 pediatric sleep studies on 3,673 unique patients conducted at Nationwide Children's Hospital between 2017 and 2019. It contains polysomnography (PSG), clinical annotations, and longitudinal clinical data.","publish_date":"2021-10-27","title":"NCH Sleep DataBank: A Large Collection of Real-world Pediatric Sleep Studies with Longitudinal Clinical Data","abstract":"In order to accelerate research on pediatric sleep and its connection to health, Nationwide Children's Hospital (NCH) and Carnegie Mellon University (CMU) introduce the NCH Sleep DataBank. This dataset has 3,984 pediatric sleep studies on 3,673 unique patients conducted at NCH in Columbus, Ohio, USA between 2017 and 2019, along with the patient's longitudinal clinical data. The published Polysomnography (PSG) contains the patient's physiological signals as well as the technician's assessment of the sleep stages and descriptions of additional irregularities.
\n\nThe novelties of this dataset include: (1) Size: Its large size is suitable for discovering new scientific insights via data mining, (2) Patient population: It explicitly focuses on pediatric patients, (3) Clinical setting: The sleep studies were gathered in the real-world clinical setting at NCH as opposed to, for example, in a controlled clinical trial, and (4) Rich set of clinical data: The accompanying 5.6 million records of clinical data are extracted from the Electronic Health Record (EHR), and are separated into encounters, medications, measurements (e.g. body mass index), diagnoses, and procedures.
\n\nThe NCH Sleep DataBank is a valuable resource for advancing automatic sleep scoring and real-time sleep disorder prediction, among many other potential scientific discoveries. Accompanying code in Python to assist users in interacting with the dataset is published on GitHub.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":2277763867374,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":["polysomnography","electronic health records","pediatrics","eeg","sleep disorders","ehr","ecg","clinical decision support","sleep study"],"source_url":"https://physionet.org/content/nch-sleep/3.1.0/"},{"public_project_uuid":"8e82d5cf-4451-4fcd-bdfc-8855462b57d0","slug":"multi-gait-posture","version":"1.0.0","core_doi":"10.13026/h6z4-fx76","version_doi":"10.13026/fyxw-n385","is_latest_version":true,"short_description":"Multimodal dataset with 166k samples for vision-based applications with a smart walker used in gait and posture rehabilitation. It is equipped with a pair of Depth cameras with data synchronized with an inertial MoCap system worn by the participant.","publish_date":"2021-11-01","title":"A multi-camera and multimodal dataset for posture and gait analysis","abstract":"Gait and posture analysis while using assisting robotic devices is of utmost importance to attain effective assistance. This work provides a multi-camera, multimodal, and detailed dataset for vision-based applications using a wheeled robotic walker equipped with a pair of affordable cameras. Depth data was acquired at 30 fps from a total of 14 healthy participants walking at 3 different gait speeds, across 3 different walking scenarios/paths at 3 different locations. Simultaneously, accurate skeleton joint data was recorded using an inertial-based commercial motion capture system that provides a reliable ground-truth for classical or novel (i.e., machine learning-based) vision-based applications. In total, the database contains approximately 166K frames of synchronized data, which amounts to 92 minutes of total recording time. This dataset may contribute to the development and evaluation of: i) classic or data-driven vision-based pose estimation algorithms; ii) applications in human detection and tracking, and movement forecasting; iii) and gait/posture metrics analysis using a rehabilitation device.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":20925680984,"compressed_storage_size":20830661677,"resource_type":"Database","access_policy":"Open","topics":["rehabilitation","inertial motion capture","deep learning","smart walker","human pose estimation","computer vision","gait and posture analysis","depth"],"source_url":"https://physionet.org/content/multi-gait-posture/1.0.0/"},{"public_project_uuid":"5a25382c-e590-4dc9-a9c6-cde4041484d0","slug":"kurias-ecg","version":"1.0","core_doi":"10.13026/ksw4-4y55","version_doi":"10.13026/kga0-0270","is_latest_version":true,"short_description":"The KURIAS-ECG database is a high-quality 12-lead ECG DB including standard vocabulary (SNOMED CT, OMOP-CDM), and ECG diagnoses of our DB are grouped into 10 diagnoses by applying the minnesota code.","publish_date":"2021-11-08","title":"KURIAS-ECG: a 12-lead electrocardiogram database with standardized diagnosis ontology","abstract":"The 12-lead electrocardiogram (ECG) is the fundamental test used to evaluate the electrophysiological state of the heart. Most previous ECG databases use the diagnosis confirmed by experts. Therefore, ECG diagnosis is not standardized and its quality is not uniform. In addition, there is a problem in that the diagnosis statements differ depending on the ECG machine. This limits the use of the database for further research, such as artificial intelligence and clinical research. Of note, modern ECG machines provide computerized rule-based ECG diagnoses comparable to physicians’ interpretations.
\n\nHere a high-quality 12-lead ECG database with standard vocabularies (SNOMED-CT and OMOP-CDM), which was transformed from a computerized ECG diagnosis of ECG machines was developed. A total of 147 ECG diagnoses were grouped into 10 categories of the Minnesota code classification. To improve the quality of the database, ECG cases of inappropriate ECG, such as poor quality, reverse arm ECG, and missing data, were removed. In addition, to minimize skew of the database, 2000 ECG cases were extracted for each Minnesota classification category. As a result of database construction, the database consisted of 20000 records for 10 categories of the Minnesota code classification from 13862 patients. The standardized database can be utilized for comprehensive research on the diagnosis of cardiac disease and the development of robust artificial intelligence technology.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":3347583596,"compressed_storage_size":2444074363,"resource_type":"Database","access_policy":"Restricted","topics":["12-lead","ecg","snomed","minnesota"],"source_url":"https://physionet.org/content/kurias-ecg/1.0/"},{"public_project_uuid":"1b59827e-2c5e-47f8-b7f5-1c9db9e04a09","slug":"auditory-eeg","version":"1.0.0","core_doi":"10.13026/s5s6-a546","version_doi":"10.13026/ps31-fc50","is_latest_version":true,"short_description":"Recording of electroencephalogram (EEG) signals with the aim to develop an EEG-based Biometric. The Data includes resting-state and auditory stimuli experiments.","publish_date":"2021-12-01","title":"Auditory evoked potential EEG-Biometric dataset","abstract":"This data set consists of over 240 two-minute EEG recordings obtained from 20 volunteers. Resting-state and auditory stimuli experiments are included in the data. The goal is to develop an EEG-based Biometric system.
\n\nThe data includes resting-state EEG signals in both cases: eyes open and eyes closed. The auditory stimuli part consists of six experiments; Three with in-ear auditory stimuli and another three with bone-conducting auditory stimuli. The three stimuli for each case are a native song, a non-native song, and neutral music.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":3072318050,"compressed_storage_size":824318816,"resource_type":"Database","access_policy":"Open","topics":["electroencephalogram","auditory stimuli","resting-state","biometric","eeg"],"source_url":"https://physionet.org/content/auditory-eeg/1.0.0/"},{"public_project_uuid":"333729b6-534d-4f3f-92f9-edd4063c9bde","slug":"treadmill-exercise-cardioresp","version":"1.0.1","core_doi":"10.13026/94yv-6145","version_doi":"10.13026/7ezk-j442","is_latest_version":true,"short_description":"Cardiorespiratory measurements of 992 treadmill maximal graded exercise tests. Heart rate, oxygen consumption, carbon dioxide generation, and pulmonary ventilation are provided.","publish_date":"2021-12-11","title":"Treadmill Maximal Exercise Tests from the Exercise Physiology and Human Performance Lab of the University of Malaga","abstract":"The present database is an ensemble of the cardiorespiratory measurements acquired during 992 treadmill maximal graded exercise tests (GET) performed in the Exercise Physiology and Human Performance Lab of the University of Malaga. Heart rate, oxygen consumption, carbon dioxide generation, and pulmonary ventilation are measured on a breath-to-breath basis along with the treadmill speed during maximal effort tests. Participants are amateur and professional athletes are of ages ranging from 10 to 63 years old. The age, height, and weight of the participants are provided, as well as the temperature and humidity during the test.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":23193152,"compressed_storage_size":6572835,"resource_type":"Database","access_policy":"Open","topics":["graded exercise testing","fitness monitoring","cardiorespiratory fitness","cardiorespiratory measures","heart rate","oxygen consumption","exercise testing","heart rate recovery"],"source_url":"https://physionet.org/content/treadmill-exercise-cardioresp/1.0.1/"},{"public_project_uuid":"e272a35b-58c6-4ead-ade8-ca1279ff790b","slug":"taichidb","version":"1.0.2","core_doi":"10.13026/9v35-xa84","version_doi":"10.13026/gq9q-rr81","is_latest_version":true,"short_description":"This project includes gait data collected with footswitches and electromyography data from subjects who walked for 10 minutes under normal conditions and for 90 seconds under dual-task conditions (walking while performing serial subtractions).","publish_date":"2021-12-14","title":"Tai Chi, Physiological Complexity, and Healthy Aging - Gait","abstract":"This dataset contains gait and electromyography (EMG) data collected during a hybrid study that included a two-arm randomized clinical trial (RCT) along with an additional observational comparison group. The RCT arm investigated the short-term effects of a Tai Chi intervention on 60 Tai Chi healthy naïve subjects, aged 50-79 years, living within the Greater Boston area, and reporting no regular Tai Chi practice within the past 5 years. Subjects were either randomized to 6 months of Tai Chi or usual care. Subjects came in for a baseline visit along with 3 month and 6 month follow-ups. The observational comparison group consisted of 27 healthy Tai Chi experts, aged 50-79 years, currently engaged in an active Tai Chi training regimen, each with at least 5 years of Tai Chi experience. The expert group was tested only at their baseline visit. The gait and EMG data were simultaneously recorded during walking at the subjects preferred speed for 10 minutes under the single-task condition and for 90 seconds under the dual-task condition. Our aim for this study was to use gait and EMG data to gain a better understanding of the mobility changes associated with Tai Chi training.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":2428065162,"compressed_storage_size":1596571588,"resource_type":"Database","access_policy":"Open","topics":["dual-task","gait","electromyography","stability"],"source_url":"https://physionet.org/content/taichidb/1.0.2/"},{"public_project_uuid":"9d711604-0b74-4385-b937-06089e1476e8","slug":"brax","version":"1.0.0","core_doi":"10.13026/083h-vh54","version_doi":"10.13026/ae9a-f727","is_latest_version":false,"short_description":"BRAX contains 24,959 chest radiography exams and 40,967 images acquired in a large general Brazilian hospital. All images have been read by trained radiologists and 14 labels were derived from Brazilian Portuguese reports using NLP.","publish_date":"2022-01-03","title":"BRAX, a Brazilian labeled chest X-ray dataset","abstract":"The Brazilian labeled chest x-ray dataset (BRAX) is an automatically labeled dataset designed to assist researchers in the validation of machine learning models. The dataset contains 24,959 chest radiography studies from patients presenting to a large general Brazilian hospital. A total of 40,967 images are available in the BRAX dataset. All images have been verified by trained radiologists and de-identified to protect patient privacy. Fourteen labels were derived from free-text radiology reports written in Brazilian Portuguese using Natural Language Processing.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":464012166939,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":["chest","artificial intelligence","x-ray","dataset"],"source_url":"https://physionet.org/content/brax/1.0.0/"},{"public_project_uuid":"edcb9adc-90df-4013-9c9a-3fd884cd0da7","slug":"circor-heart-sound","version":"1.0.0","core_doi":"10.13026/g02k-a047","version_doi":"10.13026/cy60-yn18","is_latest_version":false,"short_description":"The CirCor DigiScope Phonocardiogram is the largest multi-spot collection of heart sound signals, i.e. 5282 records collected from 1568 patients. Murmur were described based on their time, shape, pitch, grading, quality, location and spot intensity.","publish_date":"2022-01-11","title":"The CirCor DigiScope Phonocardiogram Dataset","abstract":"A total number of 5282 heart sound recordings were collected from the main four auscultation locations of 1568 patients, aged between 0.1 and 356.1 months (mean ± STD = 73.4 ± 50.3 months), with a duration between 4.8 to 80.4 seconds (mean ± STD = 22.9 ± 7.4 s), summing up to more than 312 hours of recording. For the first time, each cardiac murmur in the dataset has been annotated in detail, in terms of timing, shape, pitch, grading, quality, and location, by using a semi-supervised annotation scheme. The annotations were performed by voting between three state-of-the-art machine-based algorithms. Two independent experts later studied the consensus and mismatches between the algorithms beat-by-beat, and performed a manual annotation whenever the algorithms had disagreed. To date, the dataset is the largest pediatric heart sound dataset and it paves the way to a deeper research on the topic of auscultation-based health recommendation systems. The dataset is available online for public research on PhysioNet.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":296151557,"compressed_storage_size":238355508,"resource_type":"Database","access_policy":"Open","topics":["murmur detection","pregnant","phonocardiogram","murmur pitch","signal processing","pediatric","murmur location","murmur grading","murmur shape","murmur quality","murmur timing","murmur intensity"],"source_url":"https://physionet.org/content/circor-heart-sound/1.0.0/"},{"public_project_uuid":"801d4d20-8409-4d4c-b9d9-1d9511e2656e","slug":"circor-heart-sound","version":"1.0.1","core_doi":"10.13026/g02k-a047","version_doi":"10.13026/7bkn-d780","is_latest_version":false,"short_description":"A large collection of multi-location heart sound signals, with 5272 records collected from 1568 subjects. Heart murmurs have been annotated by a human annotator based on their time, shape, pitch, grading, quality, location and location intensity.","publish_date":"2022-01-28","title":"The CirCor DigiScope Phonocardiogram Dataset","abstract":"A total number of 5272 heart sound recordings were collected from the main four auscultation locations of 1568 subjects, aged between 0 and 21 years (mean ± STD = 6.1 ± 4.1 years), with a duration between 4.8 to 80.4 seconds (mean ± STD = 22.9 ± 7.4 s), totaling more than 33.5 hours of recording. Each cardiac murmur in the dataset has been annotated in detail by a human annotator, in terms of timing, shape, pitch, grading, quality and location. Moreover, segmentation annotations regarding the location of fundamental heart sounds (S1 and S2) in the recordings have been obtained using a semi-supervised scheme. The segmentation annotations were performed by voting between three state-of-the-art machine-based algorithms. An expert annotator later studied the consensus and mismatches between the algorithms beat-by-beat and performed a manual annotation whenever the algorithms had disagreed or were not acceptable for the expert. To date, the dataset is the largest publicly available pediatric heart sound dataset, supporting deeper research on the topic of auscultation-based health recommendation systems. The dataset is being used in the George B. Moody PhysioNet Challenge 2022 on Heart Murmur Detection from Phonocardiogram Recordings.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":586004646,"compressed_storage_size":471276295,"resource_type":"Database","access_policy":"Open","topics":["murmur grading","murmur quality","murmur intensity","murmur timing","murmur pitch","murmur detection","pediatric","murmur shape","george b moody physionet challenge 2022","signal processing","murmur location","pregnant","phonocardiogram"],"source_url":"https://physionet.org/content/circor-heart-sound/1.0.1/"},{"public_project_uuid":"f4c53180-e606-4cd6-866a-180669e30624","slug":"synthetic-mimic-iii-health-gym","version":"1.0.0","core_doi":"10.13026/qjx9-3r48","version_doi":"10.13026/p0tv-0r98","is_latest_version":true,"short_description":"This repository hosts the original Health Gym datasets of Acute Hypotension and Sepsis","publish_date":"2022-02-23","title":"Synthetic Acute Hypotension and Sepsis Datasets Based on MIMIC-III and Published as Part of the Health Gym Project","abstract":"These two synthetic datasets comprise vital signs, laboratory test results, administered fluid boluses and vasopressors for 3,910 patients with acute hypotension and for 2,164 patients with sepsis in the Intensive Care Unit (ICU). The patient cohorts were built using previously published inclusion and exclusion criteria and the data were created using Generative Adversarial Networks (GANs) and the MIMIC-III Clinical Database. The risk of identity disclosure associated with the release of these data was estimated to be very low (0.045%). The datasets were generated and published as part of the Health Gym, a project aiming to publicly distribute synthetic longitudinal health data for developing machine learning algorithms (with a particular focus on offline reinforcement learning) and for educational purposes.
\n
\nNote:
\nA PDF version of the project content is attached as a supplementary file along with this submission.
We provide an open access, high resolution and time synchronised dataset from multiple sensors worn at different body locations including Photoplethysmogram (PPG), Inertial, Pressure and ECG. The recordings are from 22 healthy subjects performing 3 physical activities. This dataset contains 66 waveform records from multi-site and multi-wavelength PPGs, sensors’ attachment pressures, sensors’ temperatures, inertial data from accelerometer and gyroscope, along with annotated ECG data for a total of 19 channels. Additionally, systolic and diastolic blood pressures, as well as blood oxygenation saturation levels (SpO2) numerics are provided.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":3102163950,"compressed_storage_size":1002385486,"resource_type":"Database","access_policy":"Open","topics":["attachment force","imu","pulse transit time","ecg","accelerometer","blood pressure","attachment pressure","finger","spo2","gyroscope","ppg","ptt"],"source_url":"https://physionet.org/content/pulse-transit-time-ppg/1.0.0/"},{"public_project_uuid":"403bb422-8f53-4d55-a54c-1e7d064abc37","slug":"eye-tracking-ecg","version":"1.0.0","core_doi":"10.13026/v5w5-n117","version_doi":"10.13026/gsr5-8b11","is_latest_version":true,"short_description":"The project aims at collecting a dataset using eye-tracking technology to understand the 12-lead electrocardiogram interpretation visual behavior for medical practitioners and students with different expertise levels.","publish_date":"2022-03-16","title":"Eye Tracking Dataset for the 12-Lead Electrocardiogram Interpretation of Medical Practitioners and Students","abstract":"This dataset contains collected eye tracking data from medical practitioners and students with different expertise levels to understand their electrocardiogram interpretation behavior. These medical practitioners encounter electrocardiograms during their daily medical practice. The end goal of the collection of this dataset was to use it for analysis to uncover insights about key best practices as well as pitfall in the ECG interpretation process. The data consists of quantitative eye tracking measurements for different eye gaze features for 63 participants. These participants belong to six expertise categories, namely: medical students, nurses, technicians, residents, fellows and finally consultants. Each one of these participants contributed to the data collection by interpreting ten different electrocardiograms. The eye tracking data includes the eye fixations count and duration, the eye gaze duration, and finally the fixations revisitations. It was collected at 60 frames per second using a Tobii Pro X2-60. Each participant was allowed a timeframe of 30 seconds to interpret each one of the ten ECGs. The collected data was processed taking into consideration two defined area of interest distributions. This enabled the extraction of metrics specifically tailored for the 12-lead electrocardiogram.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":31939552,"compressed_storage_size":27206358,"resource_type":"Database","access_policy":"Open","topics":["human vision","medical students","ecg interpretation","medical image interpretation","medical practice","ecg","medical education","human-computer interaction","eye-tracking","medical practitioners","visual expertise","electrocardiogram"],"source_url":"https://physionet.org/content/eye-tracking-ecg/1.0.0/"},{"public_project_uuid":"3b5f981a-dd89-492b-b906-a4c3a0033eff","slug":"entity-bert","version":"1.0.1","core_doi":"10.13026/vagp-v123","version_doi":"10.13026/e7kt-q579","is_latest_version":true,"short_description":"Pretraining of models with a broad representation of biomedical terminology (PubMedBERT) on MIMIC-III corpus along with or without a novel entity-centric masking strategy.","publish_date":"2022-03-17","title":"EntityBERT: BERT-based Models Pretrained on MIMIC-III with or without Entity-centric Masking Strategy for the Clinical Domain","abstract":"Transformer-based neural language models have led to breakthroughs for a variety of natural language processing (NLP) tasks. However, most models are pretrained on general domain data. We propose a methodology to produce a model focused on the clinical domain: continued pretraining of a model with a broad representation of biomedical terminology (PubMedBERT) on a clinical corpus (MIMIC-III) along with a novel entity-centric masking strategy to infuse domain knowledge in the learning process.
\n\nWe curated the MIMIC-III corpus by annotating events (including diseases/disorders, signs/symptoms, medications, anatomical sites, and procedures) and time expressions (e.g. "yesterday", "this weekend", "02/31/2028"(an example date)) with special markers. Marked events and time expressions are randomly chosen together with other words in a certain ratio to be masked for training the entity-centric mask language model. Therefore, the models are infused with clinical entity information and good for entity-related clinical NLP tasks.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1322265314,"compressed_storage_size":1225860653,"resource_type":"Model","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/entity-bert/1.0.1/"},{"public_project_uuid":"b485d764-ef47-4eaa-b73e-b6cd98f17498","slug":"calcium-imaging-sleep-state","version":"1.0.1","core_doi":"10.13026/q34j-yz38","version_doi":"10.13026/prp1-2d11","is_latest_version":true,"short_description":"Wide-field calcium imaging database that consists of annotated sleep recording collected from transgenic mice at Washington University of St Louis School of Medicine.","publish_date":"2022-03-17","title":"Wide-field calcium imaging sleep state database","abstract":"A collection of wide-field calcium imaging (WFCI) sleep and wake recordings collected from twelve transgenic mice expressing GCaMP6f in excitatory neurons. Each mouse underwent a three-hour undisturbed WFCI recording session where wake, REM (rapid eye movement) sleep and NREM (non-REM) sleep was recorded. Each WFCI recording is manually scored by sleep scoring experts in 10-second epochs as wake, NREM or REM by use of adjunct EEG/EMG. The dataset contains annotated WFCI recordings, brain mask and the Paxinos atlas used for defining the brain regions. The dataset was collected as part of a study evaluating a deep learning-based automated sleep state classification method.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":225878017293,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Open","topics":["wide-field calcium imaging","sleep state classification","sleep","machine learning","sleep staging"],"source_url":"https://physionet.org/content/calcium-imaging-sleep-state/1.0.1/"},{"public_project_uuid":"b1a5a0af-a0ff-4f4b-a6a7-41b00b71a8b1","slug":"ecg-fragment-high-risk-label","version":"1.0.0","core_doi":"10.13026/wy7j-pw36","version_doi":"10.13026/kpfg-xs25","is_latest_version":true,"short_description":"Dataset derived from the MIT-BIH Malignant Ventricular Ectopy Database.","publish_date":"2022-03-17","title":"ECG Fragment Database for the Exploration of Dangerous Arrhythmia","abstract":"The database contains a set of 2 second fragments of ECG signals with rhythm disturbances, which are grouped into separate classes according to the degree of threat to the patient's life. It is intended for practical use to develop and test the efficiency of algorithms for detecting dangerous arrhythmias for continuous monitoring systems.
\n\nThe “MIT-BIH Malignant Ventricular Ectopy Database” (MVED) was selected as the primary source of ECG records. This database contains heart rhythm disturbances necessary for research and is widely used for testing and comparative analysis of various algorithms for detecting arrhythmia threats. The signal is a modified limb lead II (MLII).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":5923520,"compressed_storage_size":4371369,"resource_type":"Database","access_policy":"Open","topics":["short ecg fragments","dangerous arrhythmias","ecg"],"source_url":"https://physionet.org/content/ecg-fragment-high-risk-label/1.0.0/"},{"public_project_uuid":"a8342d01-3043-4277-85b8-ac04aa20b68b","slug":"hmc-sleep-staging","version":"1.1","core_doi":"10.13026/t4w7-3k21","version_doi":"10.13026/t79q-fr32","is_latest_version":true,"short_description":"A collection of 151 whole-night PolySomnoGraphic (PSG) sleep recordings from the Haaglanden Medisch Centrum (HMC, The Netherlands) sleep center containing different traces of ExG activity and expert's scorings of sleep stages","publish_date":"2022-03-18","title":"Haaglanden Medisch Centrum sleep staging database","abstract":"A collection of 151 whole-night polysomnographic (PSG) sleep recordings (85 Male, 66 Female, mean Age of 53.9 ± 15.4) collected during 2018 at the Haaglanden Medisch Centrum (HMC, The Netherlands) sleep center. Patient recordings were randomly selected and include a heterogeneous population which was referred for PSG examination on the context of different sleep disorders. The dataset contains electroencephalographic (EEG), electrooculographic (EOG), chin electromyographic (EMG), and electrocardiographic (ECG) activity, as well as event annotations corresponding to scoring of sleep patterns (hypnogram) performed by sleep technicians at HMC. The dataset was collected as part of a study evaluating the generalization performance of an automatic sleep scoring algorithm across multiple heterogeneous datasets.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":16888479621,"compressed_storage_size":13876267686,"resource_type":"Database","access_policy":"Open","topics":["edf","inter-database generalization","sleep staging","psg","deep learning"],"source_url":"https://physionet.org/content/hmc-sleep-staging/1.1/"},{"public_project_uuid":"7eb09263-0d87-43b2-9dec-b9bb34ba5f62","slug":"pulse-transit-time-ppg","version":"1.1.0","core_doi":"10.13026/338y-cd69","version_doi":"10.13026/jpan-6n92","is_latest_version":true,"short_description":"Time synchronised multi-site PPG dataset for PTT including sensors’ attachment pressures, temperatures, inertial data from accelerometer and gyroscope, annotated ECG data, blood pressures, as well as blood oxygenation saturation levels (SpO2)","publish_date":"2022-03-19","title":"Pulse Transit Time PPG Dataset","abstract":"We provide an open access, high resolution and time synchronised dataset from multiple sensors worn at different body locations including Photoplethysmogram (PPG), Inertial, Pressure and ECG. The recordings are from 22 healthy subjects performing 3 physical activities. This dataset contains 66 waveform records from multi-site and multi-wavelength PPGs, sensors’ attachment pressures, sensors’ temperatures, inertial data from accelerometer and gyroscope, along with annotated ECG data for a total of 19 channels. Additionally, systolic and diastolic blood pressures, as well as blood oxygenation saturation levels (SpO2) numerics are provided.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":3102163702,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Open","topics":["finger","attachment pressure","ppg","ptt","imu","accelerometer","ecg","spo2","gyroscope","attachment force","pulse transit time","blood pressure"],"source_url":"https://physionet.org/content/pulse-transit-time-ppg/1.1.0/"},{"public_project_uuid":"7545c47d-07ed-4ea7-a9a2-ae18f14965a6","slug":"vindr-mammo","version":"1.0.0","core_doi":"10.13026/5vp3-8s83","version_doi":"10.13026/br2v-7517","is_latest_version":true,"short_description":"A large-scale benchmark dataset for computer-aided detection and diagnosis in mammography","publish_date":"2022-03-21","title":"VinDr-Mammo: A large-scale benchmark dataset for computer-aided detection and diagnosis in full-field digital mammography","abstract":"Breast cancer is one of the most prevalent types of cancer and the leading type of cancer death. Mammography is the recommended imaging modality for periodic breast cancer screening. A few datasets have been published to develop computer-aided tools for mammography analysis. However, these datasets either have a limited sample size or consist of screen-film mammography (SFM), which have been replaced by full-field digital mammography (FFDM) in clinical practices. This project introduces a large-scale full-field digital mammography dataset of 5,000 four-view exams, which are double read by experienced mammographers to provide cancer assessment and breast density following the Breast Imaging Report and Data System (BI-RADS). Breast abnormalities that require further examination are also marked by bounding rectangles.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":362752902413,"compressed_storage_size":54411087734,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/vindr-mammo/1.0.0/"},{"public_project_uuid":"6b830961-e2de-460b-b03f-b909467d5dff","slug":"vindr-pcxr","version":"1.0.0","core_doi":"10.13026/xkdz-2d35","version_doi":"10.13026/k8qc-na36","is_latest_version":true,"short_description":"An open, large-scale pediatric chest X-ray dataset that contains both lesion-level labels and image-level labels for multiple findings and diseases for interpretation of common thoracic diseases.","publish_date":"2022-03-21","title":"VinDr-PCXR: An open, large-scale pediatric chest X-ray dataset for interpretation of common thoracic diseases","abstract":"Computer-aided diagnosis systems in adult chest radiography (CXR) have recently achieved great success thanks to the availability of large-scale, annotated datasets and the advent of high-performance supervised learning algorithms. However, the development of diagnostic models for detecting and diagnosing pediatric diseases in CXR scans is undertaken due to the lack of high-quality physician-annotated datasets. To overcome this challenge, we introduce and release in this paper a new pediatric CXR dataset of 9,125 studies that were retrospectively collected from a major pediatric hospital in Vietnam between 2020-2021. Each scan was manually annotated by an experienced radiologist for the presence of 36 critical findings and 15 diseases. In particular, each abnormal finding was identified via a rectangle bounding box on the image. To the best of our knowledge, this is the first and largest pediatric CXR dataset containing lesion-level labels and image-level labels for multiple findings and diseases. For algorithm development, the dataset is divided into a training set of 7,728 and a test set of 1,397.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":36574288307,"compressed_storage_size":25100819979,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/vindr-pcxr/1.0.0/"},{"public_project_uuid":"1f98eff2-77d9-4bb6-a93f-dac04bd652ce","slug":"norwegian-athlete-ecg","version":"1.0.0","core_doi":"10.13026/863g-tm96","version_doi":"10.13026/qpjf-gk87","is_latest_version":true,"short_description":"This project contains 28 ECGs from 28 healthy elite athletes. The ECGs have been interpreted by the Marquette SL12 (version 23) algorithm and a Cardiologist using the International Criteria for ECG interpretation (2018).","publish_date":"2022-03-23","title":"Norwegian Endurance Athlete ECG Database","abstract":"The Norwegian Endurance Athlete ECG Database contains 12-lead ECG recordings from 28 elite athletes from various sports in Norway. All recordings are 10 seconds resting ECGs recorded with a General Electric (GE) MAC VUE 360 electrocardiograph. All ECGs are interpreted with both the GE Marquette SL12 algorithm (version 23 (v243)) and one cardiologist with training in interpretation of athlete's ECG. The data was collected at the University of Oslo in February and March 2020.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":3400191,"compressed_storage_size":1817922,"resource_type":"Database","access_policy":"Open","topics":["marquettesl12","ecg","ge","cardiologist","electrocardiogram","athletes"],"source_url":"https://physionet.org/content/norwegian-athlete-ecg/1.0.0/"},{"public_project_uuid":"b7624ce3-6e4a-4cc7-94f4-b809b241822c","slug":"cpap-data-canterbury","version":"1.0.1","core_doi":"10.13026/33rx-gh20","version_doi":"10.13026/xfae-vv63","is_latest_version":true,"short_description":"A pressure and flow dataset was collected from a trial of 30 adults at the University of Canterbury undergoing CPAP therapy for a variety of instructed breath rates at PEEP levels of 4cmH2O and 7cmH2O.","publish_date":"2022-03-24","title":"CPAP Pressure and Flow Data from a Local Trial of 30 Adults at the University of Canterbury","abstract":"A pressure and flow dataset for CPAP (continuous positive airway pressure) breathing obtained from 30 subjects for model-based identification of patient-specific lung mechanics using a specially designed sensor system comprising an array of differential pressure sensors (Sensirion AG SDP816-125PA) and gauge pressure sensors (NXP MPVZ4006GW7U). Relevant medical information was collected using a questionnaire, including: sex; age; weight; height; smoking history; and history of asthma. Subjects were tasked with breathing at five different rates (including passive) from very slow to very fast breathing, matched to an online pacing sound and video, at two different levels of CPAP (4 and 7 cmH2O) for between 50 and 180 seconds depending on rate. Each data set for a breathing rate comprises ~17 breaths, including rest periods between breathing rates and CPAP levels. Data with significant observed sensor error was removed during processing. A better-fitted mask would be expected to reduce error in fully capturing expiration by reducing leaks around the masks seal with the face. The sensors measure inspiration and expiration using separate unidirectional differential flow sensors, which are combined in post-processing to for complete breathing data.
","license":{"name":"Creative Commons Attribution-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":244918977,"compressed_storage_size":93709091,"resource_type":"Database","access_policy":"Open","topics":["peep","biomedical engineering","cpap","respiratory mechanics","pulmonary mechanics","respiratory modelling"],"source_url":"https://physionet.org/content/cpap-data-canterbury/1.0.1/"},{"public_project_uuid":"de3c5e74-f5a4-4944-8289-754725fe7867","slug":"semg","version":"1.0.0","core_doi":"10.13026/9q7v-6016","version_doi":"10.13026/bwvb-ht51","is_latest_version":false,"short_description":"The dataset is composed of long-lasting surface electromyographic (sEMG) signals recorded from ten muscles during ground walking of 31 young able-bodied subjects in Movement Analysis Lab, Università Politecnica delle Marche, Ancona, Italy.","publish_date":"2022-03-31","title":"Surface electromyographic signals collected during long-lasting ground walking of young able-bodied subjects","abstract":"The present dataset is composed of long-lasting (around 5 minutes) surface electromyographic (sEMG) signals recorded from 2011 and 2018 during ground walking of 31 young (20 years < age < 30 years) able-bodied subjects in the Movement Analysis Lab, Università Politecnica delle Marche, Ancona, Italy. Underweight, overweight, and obese people (body mass index, BMI < 18.5 Kg/m2 and BMI > 25 Kg/m2) and subjects affected by any pathological condition, joint pain, or undergone orthopedic surgery are not included in the present dataset. sEMG signals are acquired from the following ten different leg muscles (five per leg): gastrocnemius lateralis (GL), tibialis anterior (TA), rectus femoris (RF), hamstrings (Ham), and vastus lateralis (VL). Synchronized footswitch and electrogoniometric data are provided in order to allow users to achieve a spatial/temporal characterization of the sEMG signals. Data have been acquired in subjects walking barefoot on level ground for around 5 min at their natural speed and pace, following an eight-shaped path which includes rectilinear segments and curves. The considerable length of the signals makes this dataset very suitable for those studies where the numerosity of the data is essential, such as machine/deep learning approaches, studies for analyzing and quantifying the variability of muscle recruitment during physiological walking, and creation of reference dataset in the characterization of pathological conditions.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":447782543,"compressed_storage_size":292006036,"resource_type":"Database","access_policy":"Open","topics":["surface emg signal","walking","biomedical signals","gait analysis","muscle recruitment"],"source_url":"https://physionet.org/content/semg/1.0.0/"},{"public_project_uuid":"321b9522-c406-4f03-88dc-1b75bdb2b668","slug":"wearable-exercise-frailty","version":"1.0.0","core_doi":"10.13026/vcjs-5p09","version_doi":"10.13026/mp8k-7p27","is_latest_version":true,"short_description":"A data collection contains a wearable-based electrocardiogram and triaxial acceleration signals of 80 elderly patients with frailty after an open-heart surgery. The signals were collected while the patients were performing a series of exercise tests.","publish_date":"2022-03-31","title":"Wearable-based signals during physical exercises from patients with frailty after open-heart surgery","abstract":"This data collection contains a single-lead electrocardiogram and triaxial acceleration signals of 80 elderly frailty patients, who entered a cardiac rehabilitation program after open-heart surgery. The signals were collected using a chest-based heart rate monitor worn by the patients performing a series of exercise tests for health status assessment. The tests include veloergometry, 6-minute walk test, stair climbing test, timed up and go test, and gait analysis on a treadmill. The signals are supplemented with the demographic and clinical characteristics of the patients including the Edmonton frail scale score and quantitative results of the exercise tests. This data collection could be particularly useful in developing wearable technology-based methods for (1) assessment of gait and balance in frailty patients and (2) monitoring of heart response to daily physical stressors.
","license":{"name":"Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":2662242907,"compressed_storage_size":2468224086,"resource_type":"Database","access_policy":"Open","topics":["aging","heart rate response","frailty","accelerometer","posture","veloergometry","timed up and go","heart rate","stair climbing","physical exercise","wearable","heart rate reserve","rehabilitation","gait","electrocardiogram","walk test","balance"],"source_url":"https://physionet.org/content/wearable-exercise-frailty/1.0.0/"},{"public_project_uuid":"1e83ebe3-30a0-4d60-82bb-ae7e293f66c1","slug":"rumednli-russian-inference","version":"1.0.0","core_doi":"10.13026/8nyz-cn98","version_doi":"10.13026/gxzd-cf80","is_latest_version":true,"short_description":"RuMedNLI is the full counterpart dataset of MedNLI in Russian language.","publish_date":"2022-04-01","title":"RuMedNLI: A Russian Natural Language Inference Dataset For The Clinical Domain","abstract":"There is a shortage of text medical resources for the Russian language. This is a substantial obstacle in state-of-the-art NLP deep learning models research and development. To mitigate this issue we translated the MedNLI data from English to Russian. The RuMedNLI task became part of the RuMedBench: Russian Medical Language Understanding Benchmark.
\n\nMedNLI is a dataset based on MIMIC-III records and annotated by doctors performing a natural language inference task (NLI) grounded in the medical history of patients. RuMedNLI is the full counterpart dataset of MedNLI in the Russian language.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":17971209,"compressed_storage_size":2273839,"resource_type":"Database","access_policy":"Credentialed","topics":["natural language inference","recognizing textual entailment","russian language"],"source_url":"https://physionet.org/content/rumednli-russian-inference/1.0.0/"},{"public_project_uuid":"05617248-aae7-4736-bec9-8a9dbdf58e72","slug":"icentia11k-continuous-ecg","version":"1.0","core_doi":"10.13026/dxjr-s643","version_doi":"10.13026/kk0v-r952","is_latest_version":true,"short_description":"This is a dataset of continuous raw electrocardiogram (ECG) signals for representation learning containing 11 thousand patients and 2 billion labelled beats.","publish_date":"2022-04-12","title":"Icentia11k Single Lead Continuous Raw Electrocardiogram Dataset","abstract":"This is a dataset of continuous raw electrocardiogram (ECG) signals containing 11 thousand patients and 2 billion labelled beats. The signals were recorded with a 16-bit resolution at 250Hz with a fixed chest mounted single lead probe for up to 2 weeks. The average age of the patient is 62.2±17.4 years. 20 technologists annotated each beat's type (Normal, Premature Atrial Contraction, Premature Ventricular contraction) and rhythm (Normal Sinusal Rhythm, Atrial Fibrillation, Atrial Flutter).
","license":{"name":"Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":1159619029738,"compressed_storage_size":202154613899,"resource_type":"Database","access_policy":"Open","topics":["ecg","representation learning"],"source_url":"https://physionet.org/content/icentia11k-continuous-ecg/1.0/"},{"public_project_uuid":"d77222f5-e0d2-4005-b152-0ddaa068f597","slug":"drugehrqa","version":"1.0.0","core_doi":"10.13026/chym-n302","version_doi":"10.13026/a849-cd06","is_latest_version":true,"short_description":"DrugEHRQA is a QA dataset containing question-answers from MIMIC-III tables and discharge summaries.","publish_date":"2022-04-12","title":"DrugEHRQA: A Question Answering Dataset on Structured and Unstructured Electronic Health Records For Medicine Related Queries","abstract":"Electronic Health Records (EHR) contain patient records, stored in structured tables as well as unstructured clinical notes. The information in structured and unstructured EHR records is not strictly disjoint: information may be duplicated, contradictory, or provide additional context between these sources. This presents a rich opportunity to study question answering (QA) models that combine reasoning over both structured and unstructured data. This work presents the first question answering (QA) dataset (DrugEHRQA) containing question-answer pairs from both structured tables and unstructured notes from MIMIC-III, a publicly available Electronic Health Record (EHR). We are releasing a QA dataset over MIMIC-III tables through PhysioNet, containing 41,417 triplets of natural language questions, its corresponding SQL query and the answer retrieved from MIMIC-III tables. We also generated a QA dataset on the unstructured clinical notes of MIMIC-III which can be found in the n2c2 repository. Both these datasets are combined to generate a multimodal QA dataset (DrugEHRQA), which contains question-answers from both structured and unstructured data of MIMIC-III. The DrugEHRQA dataset has medication-related queries, containing over 70,000 question-answer pairs. Our goal is to provide a benchmark dataset for multi-modal QA systems, and to open up new avenues of research in improving question answering over EHR structured data by using context from unstructured clinical data.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":20569544,"compressed_storage_size":2122430,"resource_type":"Database","access_policy":"Credentialed","topics":["question-answer","qa"],"source_url":"https://physionet.org/content/drugehrqa/1.0.0/"},{"public_project_uuid":"9f1826d8-abe4-48f4-ac2d-aa8db8c75be9","slug":"circor-heart-sound","version":"1.0.2","core_doi":"10.13026/g02k-a047","version_doi":"10.13026/qcpr-xg17","is_latest_version":false,"short_description":"A large collection of multi-location heart sound signals, with 5272 records collected from 1568 subjects. Heart murmurs have been annotated by a human annotator based on their time, shape, pitch, grading, quality, location and location intensity.","publish_date":"2022-04-30","title":"The CirCor DigiScope Phonocardiogram Dataset","abstract":"A total number of 5272 heart sound recordings were collected from the main four auscultation locations of 1568 subjects, aged between 0 and 21 years (mean ± STD = 6.1 ± 4.3 years), with a duration between 4.8 to 80.4 seconds (mean ± STD = 22.9 ± 7.4 s), totaling more than 33.5 hours of recording. Each cardiac murmur in the dataset has been annotated in detail by a human annotator, in terms of timing, shape, pitch, grading, quality and location. Moreover, segmentation annotations regarding the location of fundamental heart sounds (S1 and S2) in the recordings have been obtained using a semi-supervised scheme. The segmentation annotations were performed by voting between three state-of-the-art machine-based algorithms. An expert annotator later studied the consensus and mismatches between the algorithms beat-by-beat and performed a manual annotation whenever the algorithms had disagreed or were not acceptable for the expert. To date, the dataset is the largest publicly available pediatric heart sound dataset, supporting deeper research on the topic of auscultation-based health recommendation systems. The dataset is being used in the George B. Moody PhysioNet Challenge 2022 on Heart Murmur Detection from Phonocardiogram Recordings.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":586007882,"compressed_storage_size":471283843,"resource_type":"Database","access_policy":"Open","topics":["signal processing","murmur quality","murmur location","phonocardiogram","murmur grading","murmur intensity","pediatric","murmur shape","george b moody physionet challenge 2022","murmur timing","murmur pitch","pregnant","murmur detection"],"source_url":"https://physionet.org/content/circor-heart-sound/1.0.2/"},{"public_project_uuid":"dd7ad4c8-5248-45ea-8168-efe023e9817d","slug":"ecg-ppg-simulator-arrhythmia","version":"1.3.1","core_doi":"10.13026/xf56-pm38","version_doi":"10.13026/s32e-sv15","is_latest_version":true,"short_description":"A model is capable of simulating sinus rhythm, atrial fibrillation and ectopic beats in ECGs and PPGs as well as extreme bradycardia and ventricular tachycardia in PPGs. Different types of noises and artifacts can also be added to the waveforms.","publish_date":"2022-05-02","title":"Model for Simulating ECG and PPG Signals with Arrhythmia Episodes","abstract":"The simulation of electrocardiographic (ECG) and photoplethysmographic (PPG) signals with cardiac arrhythmias is important when developing and testing arrhythmia detectors. This software is capable of simulating sinus rhythm, episodes of atrial fibrillation and atrial premature beats in ECGs and PPGs as well as extreme bradycardia and ventricular tachycardia in PPGs. Different types of noise and artifacts can be added to simulated signals to improve realism and mimic the signals acquired in free-living conditions. Two versions are provided: a binary version for Windows 10 and Matlab source code.
","license":{"name":"GNU General Public License version 3"},"dua":null,"main_storage_size":5137673018,"compressed_storage_size":4938664477,"resource_type":"Software","access_policy":"Open","topics":["arrhythmia","tachycardia","atrial fibrillation","detection","noise","motion artifacts","ppg","ecg","simulation","bradycardia"],"source_url":"https://physionet.org/content/ecg-ppg-simulator-arrhythmia/1.3.1/"},{"public_project_uuid":"36ed4ef4-06d8-4c86-ab35-41ec834f4c44","slug":"circor-heart-sound","version":"1.0.3","core_doi":"10.13026/g02k-a047","version_doi":"10.13026/tshs-mw03","is_latest_version":true,"short_description":"A large collection of multi-location heart sound signals, with 5272 records collected from 1568 subjects. Heart murmurs have been annotated by a human annotator based on their time, shape, pitch, grading, quality, location and location intensity.","publish_date":"2022-05-10","title":"The CirCor DigiScope Phonocardiogram Dataset","abstract":"A total number of 5272 heart sound recordings were collected from the main four auscultation locations of 1568 subjects, aged between 0 and 21 years (mean ± STD = 6.1 ± 4.3 years), with a duration between 4.8 to 80.4 seconds (mean ± STD = 22.9 ± 7.4 s), totaling more than 33.5 hours of recording. Each cardiac murmur in the dataset has been annotated in detail by a human annotator, in terms of timing, shape, pitch, grading, quality and location. Moreover, segmentation annotations regarding the location of fundamental heart sounds (S1 and S2) in the recordings have been obtained using a semi-supervised scheme. The segmentation annotations were performed by voting between three state-of-the-art machine-based algorithms. An expert annotator later studied the consensus and mismatches between the algorithms beat-by-beat and performed a manual annotation whenever the algorithms had disagreed or were not acceptable for the expert. To date, the dataset is the largest publicly available pediatric heart sound dataset, supporting deeper research on the topic of auscultation-based health recommendation systems. The dataset is being used in the George B. Moody PhysioNet Challenge 2022 on Heart Murmur Detection from Phonocardiogram Recordings.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":586007899,"compressed_storage_size":471283852,"resource_type":"Database","access_policy":"Open","topics":["signal processing","murmur pitch","george b moody physionet challenge 2022","murmur grading","murmur location","murmur timing","phonocardiogram","pregnant","murmur shape","pediatric","murmur detection","murmur intensity","murmur quality"],"source_url":"https://physionet.org/content/circor-heart-sound/1.0.3/"},{"public_project_uuid":"4103ebb1-076f-4393-993f-33618b402487","slug":"brateca","version":"1.0","core_doi":"10.13026/v8a6-mr20","version_doi":"10.13026/cmab-j041","is_latest_version":false,"short_description":"Brazilian clinical dataset containing over 70,000 admissions from 10 hospitals in two Brazilian states.","publish_date":"2022-05-13","title":"BRATECA (Brazilian Tertiary Care Dataset): a Clinical Information Dataset for the Portuguese Language","abstract":"Computational medicine research requires clinical data for training and testing purposes, so the development of datasets composed of real hospital data is of utmost importance in this field. Most such data collections are in the English language, were collected in anglophone countries, and do not reflect other clinical realities, which increases the importance of such national datasets for projects that hope to positively impact public health. This paper presents a new Brazilian Clinical Dataset containing over 70,000 admissions from 10 hospitals in two Brazilian states, composed of a sum total of over 2.5 million free-text clinical notes alongside data pertaining to patient descriptors, prescription information, and exam results. This data was collected, organized, deidentified, and is being distributed via credentialed access for the use of the research community. In the course of presenting the new dataset, we explore the new dataset’s structure, population, and potential benefits for use in clinical AI tasks.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1763370863,"compressed_storage_size":103110672,"resource_type":"Database","access_policy":"Credentialed","topics":["exams","natural language processing","tertiary care","prescriptions","clinical notes"],"source_url":"https://physionet.org/content/brateca/1.0/"},{"public_project_uuid":"282d487a-02cc-4ae8-a108-7d4a407c8fc9","slug":"ms-cxr","version":"0.1","core_doi":"10.13026/ms5s-2a34","version_doi":"10.13026/b90j-vb87","is_latest_version":false,"short_description":"MS-CXR is a new dataset containing 1162 Chest X-ray bounding box labels paired with radiology text descriptions, annotated and verified by two board-certified radiologists.","publish_date":"2022-05-16","title":"MS-CXR: Making the Most of Text Semantics to Improve Biomedical Vision-Language Processing","abstract":"We release a new dataset, MS-CXR, with locally-aligned phrase grounding annotations by board-certified radiologists to facilitate the study of complex semantic modelling in biomedical vision–language processing. The MS-CXR dataset provides 1162 image–sentence pairs of bounding boxes and corresponding phrases, collected across eight different cardiopulmonary radiological findings, with an approximately equal number of pairs for each finding. This dataset complements the existing MIMIC-CXR v.2 dataset and comprises: 1. Reviewed and edited bounding boxes and phrases (1026 pairs of bounding box/sentence); and 2. Manual bounding box labels from scratch (136 pairs of bounding box/sentence).
\n\nThis large, well-balanced phrase grounding benchmark dataset contains carefully curated image regions annotated with descriptions of eight radiology findings, as verified by radiologists. Unlike existing chest X-ray benchmarks, this challenging phrase grounding task evaluates joint, local image-text reasoning while requiring real-world language understanding, e.g. to parse domain-specific location references, complex negations, and bias in reporting style. This data accompany work showing that principled textual semantic modelling can improve contrastive learning in self-supervised vision–language processing.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":3555169,"compressed_storage_size":406921,"resource_type":"Database","access_policy":"Credentialed","topics":["chest x-ray","vision-language processing"],"source_url":"https://physionet.org/content/ms-cxr/0.1/"},{"public_project_uuid":"5408449b-1e23-479a-986d-70e1b728870c","slug":"kinecal","version":"1.0.0","core_doi":"10.13026/8bsb-ch59","version_doi":"10.13026/pq4z-4g10","is_latest_version":false,"short_description":"A dataset for balance falls-risk assessment and balance impairment analysis","publish_date":"2022-05-16","title":"KINECAL","abstract":"The field of human action recognition has made great strides in recent years, much helped by the availability of a wide variety of datasets that use Kinect to record human movement. Conversely, progress towards using Kinect in clinical practice has been hampered by the lack of appropriate data. In particular, datasets that contain clinically significant movements and appropriate metadata. This paper proposes a dataset to address this issue, namely KINECAL. It contains the recordings of 90 individuals carrying out 11 movements, commonly used in the clinical assessment of balance. The dataset contains relevant metadata, including clinical labelling, falls history and postural sway metrics. In addition, the recordings were made in both the lab and in informal settings.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":192423497685,"compressed_storage_size":92063954192,"resource_type":"Database","access_policy":"Restricted","topics":["posturography","postural sway","balance","age-related changes","clinical tests","falls-risk"],"source_url":"https://physionet.org/content/kinecal/1.0.0/"},{"public_project_uuid":"d5b89e11-00fd-4464-90d5-18fed281f1cf","slug":"heart-failure-zigong","version":"1.3","core_doi":"10.13026/wd7p-qy77","version_doi":"10.13026/5m60-vs44","is_latest_version":true,"short_description":"The new version added beta blockers in the dat_md.csv file. Dataset comprising hospital-level data on patients who were admitted with heart failure to Zigong Fourth People’s Hospital, Sichuan, China between 2016 and 2019.","publish_date":"2022-05-22","title":"Hospitalized patients with heart failure: integrating electronic healthcare records and external outcome data","abstract":"Heart failure is a common reason for hospitalization in the elderly and it is associated with significant mortality and morbidity. To facilitate epidemiological studies of heart failure, there is a need for high quality datasets to be made available to researchers. While several heart failure datasets have been established in Western countries, there is a paucity of data available from China. Understanding differences in patient populations and healthcare systems between China and other countries is important in providing optimal care. To help address this issue, we created a retrospective heart failure dataset using electronic health data collected from patients who were admitted to a hospital in Sichuan, China between 2016 and 2019. The dataset includes 168 variables for 2,008 patients with heart failure.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":2163749,"compressed_storage_size":499031,"resource_type":"Database","access_policy":"Restricted","topics":["heart failure","electronic health record","china"],"source_url":"https://physionet.org/content/heart-failure-zigong/1.3/"},{"public_project_uuid":"c5a067f0-b247-4fc3-98c4-d506a9221573","slug":"wearable-exam-stress","version":"1.0.0","core_doi":"10.13026/wn66-m424","version_doi":"10.13026/kvkb-aj90","is_latest_version":true,"short_description":"The data contains electrodermal activity, heart rate, blood volume pulse, skin surface temperature, inter beat interval and accelerometer data recorded during three exam sessions (midterm 1, midterm 2 and finals) as well as their corresponding grades","publish_date":"2022-05-26","title":"A Wearable Exam Stress Dataset for Predicting Cognitive Performance in Real-World Settings","abstract":"Preliminary datasets containing skin conductance recordings where experiments involve artificial stimuli are available online. However, recordings of a longer duration acquired during real-world cognitive stressors are lacking significantly. This dataset seeks to bridge the gap between a body of knowledge concerning the human stress response primarily built up on laboratory acquired data, and how people’s skin conductance fluctuates in reality when they undergo a stressful exam experience. The data contains electrodermal activity, heart rate, blood volume pulse, skin surface temperature, inter beat interval and accelerometer data recorded during three exam sessions (midterm 1, midterm 2 and finals) as well as their corresponding grades. The purpose of this dataset is to gauge the influence of stress on exam performance.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":86038611,"compressed_storage_size":85968624,"resource_type":"Database","access_policy":"Open","topics":["stress"],"source_url":"https://physionet.org/content/wearable-exam-stress/1.0.0/"},{"public_project_uuid":"f412bc1f-4817-4486-9c18-69ce0231cc1f","slug":"bp-graphene-bioimpedance","version":"1.0.0","core_doi":"10.13026/qcc8-n557","version_doi":"10.13026/ce62-pc98","is_latest_version":true,"short_description":"Cuffless blood pressure data repository that includes raw time data for 4-channel Bioimpedance signals using Graphene Tattoos from the wrist with synchronized continuous blood pressure and PPG signals from 7 subjects","publish_date":"2022-06-05","title":"Continuous Cuffless Monitoring of Arterial Blood Pressure via Graphene Bioimpedance Tattoos","abstract":"\nContinuous monitoring of arterial blood pressure (BP) in non-clinical (ambulatory) settings is essential for a proper understanding of numerous health conditions. While conventional ambulatory BP devices exist, they are uncomfortable, bulky, and intrusive. In this work, we introduce a dataset of BP measurement using atomically-thin graphene electronic tattoos (GETs) and bioimpedance (Bio-Z) as measurement modality. The Bio-Z is measured from radial and ulnar arteries on the wrist. Afterward, a machine learning regression model is employed to relate the recorded bioimpedance to BP, yielding an effective beat-to-beat BP monitoring platform. In this database, we are sharing the raw time data for the 4-channel GET Bioimpedance signals with BP and PPG signals. The GETs are used to monitor arterial blood pressure from N=7 individuals for 300+ minutes each. A medical-grade BP monitoring device, Finapres NOVA, was used to measure the subject's control BP during the experiments. On average, each participant underwent 4±1 hours of continuous BP monitoring. Altogether, we performed on average ~2500±600 BP measurements per participant, which resulted in a total of 18667 datapoints.
\n","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":2436679577,"compressed_storage_size":787660833,"resource_type":"Database","access_policy":"Open","topics":["graphene","bioimpedance","ppg","blood pressure"],"source_url":"https://physionet.org/content/bp-graphene-bioimpedance/1.0.0/"},{"public_project_uuid":"8072adfa-c204-42b0-a0e4-1a7d9d7f49ff","slug":"mimic-iv-fhir-demo","version":"2.0","core_doi":"10.13026/vphg-y548","version_doi":"10.13026/2f5g-dh02","is_latest_version":false,"short_description":"MIMIC-IV-on-FHIR is a hundred patient demo of MIMIC-IV v2.0 in the Fast Healthcare Interoperability Resources(FHIR) format. MIMIC-IV-on-FHIR provides implementers with a real-world FHIR datastore to aid in FHIR research and development.","publish_date":"2022-06-07","title":"MIMIC-IV Clinical Database Demo on FHIR","abstract":"Interoperability of healthcare data has become increasingly important given the increase in deployment of data driven algorithms in clinical settings. The Fast Healthcare Interoperability Resources (FHIR) standard has emerged as a promising mechanism to share healthcare data across vendors in real-time and batch settings. Real-world datasets available in FHIR would accelerate research and development of data-driven algorithms. Existing datasets in FHIR are primarily synthetic, and cover a limited number of resources. To address this gap, we have reformatted the Medical Information Mart for Intensive Care (MIMIC)-IV Clinical Database Demo into FHIR. The MIMIC clinical databases have received wide adoption and the constituent data are understood by the community. As much as possible, we adhered to the base resources with minimal extensions. Alongside the dataset, we publish openly available code allowing researchers to quickly build upon our work. Translating MIMIC-IV into FHIR provides a benchmark dataset for institutions to experiment with FHIR based tools, and we hope this resource supports adoption and use of FHIR.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":918715918,"compressed_storage_size":121863908,"resource_type":"Database","access_policy":"Open","topics":["mimic","fhir","electronic health records"],"source_url":"https://physionet.org/content/mimic-iv-fhir-demo/2.0/"},{"public_project_uuid":"67d082bb-133c-4c56-b94e-1582246fbc68","slug":"mimiciv","version":"2.0","core_doi":"10.13026/07hj-2a80","version_doi":"10.13026/7vcr-e114","is_latest_version":false,"short_description":"Database of patients admitted to the Beth Israel Deaconess Medical Center","publish_date":"2022-06-13","title":"MIMIC-IV","abstract":"Retrospectively collected medical data has the opportunity to improve patient care through knowledge discovery and algorithm development. Broad reuse of medical data is desirable for the greatest public good, but data sharing must be done in a manner which protects patient privacy. The Medical Information Mart for Intensive Care (MIMIC)-III database provided critical care data for over 40,000 patients admitted to intensive care units at the Beth Israel Deaconess Medical Center (BIDMC). Importantly, MIMIC-III was deidentified, and patient identifiers were removed according to the Health Insurance Portability and Accountability Act (HIPAA) Safe Harbor provision. MIMIC-III has been integral in driving large amounts of research in clinical informatics, epidemiology, and machine learning. Here we present MIMIC-IV, an update to MIMIC-III, which incorporates contemporary data and improves on numerous aspects of MIMIC-III. MIMIC-IV adopts a modular approach to data organization, highlighting data provenance and facilitating both individual and combined use of disparate data sources. MIMIC-IV is intended to carry on the success of MIMIC-III and support a broad set of applications within healthcare.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":7925729028,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":["machine learning","critical care","intensive care unit","mimic"],"source_url":"https://physionet.org/content/mimiciv/2.0/"},{"public_project_uuid":"5e523a0f-91f0-4fc7-8077-2180e8efa3d8","slug":"brax","version":"1.1.0","core_doi":"10.13026/083h-vh54","version_doi":"10.13026/grwk-yh18","is_latest_version":true,"short_description":"BRAX contains 24,959 chest radiography exams and 40,967 images acquired in a large general Brazilian hospital. All images have been read by trained radiologists and 14 labels were derived from Brazilian Portuguese reports using NLP.","publish_date":"2022-06-17","title":"BRAX, a Brazilian labeled chest X-ray dataset","abstract":"The Brazilian labeled chest x-ray dataset (BRAX) is an automatically labeled dataset designed to assist researchers in the validation of machine learning models. The dataset contains 24,959 chest radiography studies from patients presenting to a large general Brazilian hospital. A total of 40,967 images are available in the BRAX dataset. All images have been verified by trained radiologists and de-identified to protect patient privacy. Fourteen labels were derived from free-text radiology reports written in Brazilian Portuguese using Natural Language Processing.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":464012248873,"compressed_storage_size":325523721967,"resource_type":"Database","access_policy":"Credentialed","topics":["chest","x-ray","artificial intelligence","dataset"],"source_url":"https://physionet.org/content/brax/1.1.0/"},{"public_project_uuid":"e0396795-45cf-413e-8a9e-5ff53725c705","slug":"wfdb","version":"10.7.0","core_doi":"10.13026/zx10-se90","version_doi":"10.13026/gjvw-1m31","is_latest_version":true,"short_description":"The original Waveform Database Software (WFDB) Package, in C.","publish_date":"2022-06-20","title":"WFDB Software Package","abstract":"Effective processing and analysis of physiological data requires specialized software. We have developed a large collection of such software over the past thirty years, and much of it is contained within the WFDB (Waveform Database) Software Package. The WFDB Software Package comprises over 70 applications for signal processing and automated analysis. A comprehensive collection of documentation, including tutorials and reference manuals, is also included in the package. The package is written in highly portable C and can be used on all popular platforms, including GNU/Linux, MacOS, MS-Windows, and all versions of Unix. The package has also been ported to other popular languages, including Matlab and Python.
","license":{"name":"GNU General Public License version 3"},"dua":null,"main_storage_size":7364036,"compressed_storage_size":3967872,"resource_type":"Software","access_policy":"Open","topics":["waveform","wfdb"],"source_url":"https://physionet.org/content/wfdb/10.7.0/"},{"public_project_uuid":"310e3bee-ce41-44ad-9c82-7e6be6b4e51b","slug":"mimic-iv-demo","version":"1.0","core_doi":"10.13026/ng9m-3n32","version_doi":"10.13026/jwtp-v091","is_latest_version":false,"short_description":"An openly available subset of patients in the MIMIC-IV database.","publish_date":"2022-06-22","title":"MIMIC-IV Clinical Database Demo","abstract":"The Medical Information Mart for Intensive Care (MIMIC)-IV database is comprised of deidentified electronic health records for patients admitted to the Beth Israel Deaconess Medical Center. Access to MIMIC-IV is limited to credentialed users. Here, we have provided an openly-available demo of MIMIC-IV containing a subset of 100 patients. The dataset includes similar content to MIMIC-IV, but excludes free-text clinical notes. The demo may be useful for running workshops and for assessing whether the MIMIC-IV is appropriate for a study before making an access request.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":14912072,"compressed_storage_size":14877490,"resource_type":"Database","access_policy":"Open","topics":["electronic health record","mimic","critical care"],"source_url":"https://physionet.org/content/mimic-iv-demo/1.0/"},{"public_project_uuid":"cfb6f76f-b4ab-430d-94c9-843c12f7eba8","slug":"mimic-iv-ed","version":"2.0","core_doi":"10.13026/1cjn-2370","version_doi":"10.13026/as7t-c445","is_latest_version":false,"short_description":"A large database of emergency department admissions.","publish_date":"2022-06-27","title":"MIMIC-IV-ED","abstract":"MIMIC-ED is a large, freely available database of emergency department (ED) admissions at the Beth Israel Deaconess Medical Center between 2011 and 2019. As of MIMIC-ED v1.0, the database contains 448,972 ED stays. Vital signs, triage information, medication reconciliation, medication administration, and discharge diagnoses are available. All data are deidentified to comply with the Health Information Portability and Accountability Act (HIPAA) Safe Harbor provision. MIMIC-ED is intended to support a diverse range of education initiatives and research studies.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":128429395,"compressed_storage_size":128381565,"resource_type":"Database","access_policy":"Credentialed","topics":["emergency","mimic","mimic-iv","ed"],"source_url":"https://physionet.org/content/mimic-iv-ed/2.0/"},{"public_project_uuid":"c0bcc9e3-889b-410e-ae38-fe78e21c2b9b","slug":"actes-cycloergometer-exercise","version":"1.0.0","core_doi":"10.13026/x1f6-1q24","version_doi":"10.13026/2qs3-kh43","is_latest_version":true,"short_description":"Cardiorespiratory measurements acquired during 18 exercise tests performed at French West Indies University","publish_date":"2022-06-30","title":"Cardiorespiratory measurement from graded cycloergometer exercise testing","abstract":"The present database is an ensemble of cardiorespiratory measurements acquired during 18 cycloergometer maximal graded exercise tests performed in the "Adaptation, Climat Tropical, Exercice et Santé" (ACTES) laboratory of the French West Indies University. Data correspond to heart R-wave to R-wave intervals, Oxygen consumption, and mechanical power output on a beat-to-beat basis recorded all along the graded exercise test sessions. Participants are 18 teenagers athletes (15.2±2 years) from the Regional Physical and Sports Education Centre (CREPS) of French West Indies (Guadeloupe, France), belonging to a regional division of sprint kayak and triathlon or a national division of fencing. Clinical data of the participants (i.e. age, Weight, Height, Power at ventilatory threshold 1 and 2, and sport discipline) are provided in a separate file.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1725970,"compressed_storage_size":449874,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/actes-cycloergometer-exercise/1.0.0/"},{"public_project_uuid":"d0068477-1603-4ef2-a67d-b67b88a6bb4d","slug":"icu-infection-zigong-fourth","version":"1.1","core_doi":"10.13026/cgzc-vp58","version_doi":"10.13026/xpt9-z726","is_latest_version":true,"short_description":"Routinely collected data from critical care units at Zigong Fourth People’s Hospital, Sichuan, China for patients admitted between January 2019 and December 2020 Missing information on temperature are updated in the new version.","publish_date":"2022-06-30","title":"Critical care database comprising patients with infection at Zigong Fourth People's Hospital","abstract":"Patients treated in the intensive care unit (ICU) are closely monitored and receive intensive treatment. Such aggressive monitoring and treatment will generate high-granularity data from both electronic healthcare records and nursing chart. These data not only provide infrastructure for daily clinical practice, but also can help to inform clinical researches. It is technically challenging to integrate and cleanse medical data from a variety of sources. Although there are several open access critical care databases from western countries, there is a lack of this kind of database for Chinese adult patients. We established a critical care database involving patients with infection. A large proportion of these patients have sepsis and/or septic shock. High-granularity data comprising laboratory findings, baseline characteristics, medications, ICD code, nursing charts and follow-up results were integrated to generate a comprehensive database. The database can be utilized for a variety of clinical studies.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":78717813,"compressed_storage_size":78656154,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/icu-infection-zigong-fourth/1.1/"},{"public_project_uuid":"2f6a4c08-8557-4ea8-b452-2add434e5c59","slug":"mimic-iv-ecg-demo","version":"0.1","core_doi":"10.13026/cnhy-ra15","version_doi":"10.13026/4eqn-kt76","is_latest_version":true,"short_description":"The MIMIC-IV ECG Demo module contains 659 diagnostic electrocardiograms across 92 unique patients. These 92 patients overlap with the patients from the MIMIC-IV Clinical Demo and are also part of the MIMIC-IV Clinical Database.","publish_date":"2022-06-30","title":"MIMIC-IV-ECG Demo - Diagnostic Electrocardiogram Matched Subset Demo","abstract":"The MIMIC-IV-ECG Demo module contains 659 diagnostic electrocardiograms across 92 unique patients. These 92 patients overlap with the patients from the MIMIC-IV Clinical Demo and are also part of the MIMIC-IV Clinical Database. These diagnostic ECGs use 12 leads and are 10 seconds in length. They are sampled at 500 Hz. This open access demo allows researchers to assess whether the MIMIC-IV-ECG module is appropriate for their research and facilitates demonstration of these diagnostic ECGs.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":79748881,"compressed_storage_size":31043363,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/mimic-iv-ecg-demo/0.1/"},{"public_project_uuid":"8e5a185b-c211-480e-a704-2bf37e5c2f1c","slug":"mimicel-ed","version":"1.0.0","core_doi":"10.13026/j8yc-gg94","version_doi":"10.13026/64aq-jw11","is_latest_version":false,"short_description":"MIMIC-IV Event Log for Emergency Department","publish_date":"2022-07-07","title":"MIMICEL: MIMIC-IV Event Log for Emergency Department","abstract":"In this work, we extract an event log from the MIMIC-IV-ED dataset by adopting a well-established event log generation methodology, and we name this event log MIMICEL. The data tables in the MIMIC-IV-ED dataset relate to each other based on the existing relational database schema, and each table records the individual activities of patients along their journey in the emergency department (ED). While the data tables in the MIMIC-IV-ED dataset catch snapshots of a patient journey in the ED, the extracted event log MIMICEL aims to capture an end-to-end patient journey process. This will enable us to analyse the existing patient flows, thereby improving the efficiency of an ED process.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":9138907841,"compressed_storage_size":503252653,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimicel-ed/1.0.0/"},{"public_project_uuid":"d3f1684a-4d45-4340-bed8-87a783007387","slug":"mimic4wdb","version":"0.1.0","core_doi":"10.13026/6269-ws81","version_doi":"10.13026/a2mw-f949","is_latest_version":true,"short_description":"The MIMIC-IV Waveform Database collects physiological signals and measurements from ICU bedside monitors. Coupled with the clinical information available in MIMIC-IV, it provides a detailed view into the physiology of critically ill patients.","publish_date":"2022-07-10","title":"MIMIC-IV Waveform Database","abstract":"The MIMIC-IV Waveform Database is a large collection of physiological signals and measurements from patients in intensive care units, including electrocardiograms, photoplethysmograms, respiration, invasive and non-invasive blood pressure, and more. These measurements and signals are obtained directly from the bedside monitor, and provide a detailed view into the physiology of critically ill patients.
\n\nCombining this database with the clinical information found in MIMIC-IV provides a broad, cross-sectional example of the data available to caregivers in a modern ICU. We hope that this database will provide a foundation for future improvements to monitoring technology as well as data-driven diagnosis and treatment.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":13732709669,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/mimic4wdb/0.1.0/"},{"public_project_uuid":"719e7f0a-9f56-4396-b0ef-0a83d7d39953","slug":"grabmyo","version":"1.0.0","core_doi":"10.13026/cadk-kw71","version_doi":"10.13026/rtfc-np50","is_latest_version":false,"short_description":"We present the Gesture Recognition and Biometrics ElectroMyogram (GRABMyo) dataset, an open-access dataset of electromyogram (EMG) recordings collected from the wrist and forearm muscles while performing hand gestures.","publish_date":"2022-07-13","title":"Gesture Recognition and Biometrics ElectroMyogram (GRABMyo)","abstract":"We present the Gesture Recognition and Biometrics ElectroMyogram (GRABMyo) dataset, an open-access dataset of electromyogram (EMG) recordings collected from the wrist and forearm muscles while performing hand gestures. Data were collected from 43 healthy participants (age range: 24-35 years) on three different days while performing 16 hand and finger gestures in identical experimental sessions on each day. The GRABMyo dataset can be used for research on: 1) EMG-based biometrics for personal identification and verification, and 2) EMG-based gesture recognition for neurorehabilitation and home applications. The large sample size will provide sufficient power for establishing results, specifically for applications such as biometrics. Further, it will be useful for subject-independent applications such as generalized classification models for gesture recognition. The multiday recordings will provide results that will be consistent over a longer duration, crucial for the reliability of EMG-based wearable devices.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":10114185087,"compressed_storage_size":9819168913,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/grabmyo/1.0.0/"},{"public_project_uuid":"a0705ca6-66c9-42c5-91a6-4bf4a47b844a","slug":"brateca","version":"1.1","core_doi":"10.13026/v8a6-mr20","version_doi":"10.13026/ay8n-qf21","is_latest_version":true,"short_description":"Brazilian clinical dataset containing over 70,000 admissions from 10 hospitals in two Brazilian states.","publish_date":"2022-07-14","title":"BRATECA (Brazilian Tertiary Care Dataset): a Clinical Information Dataset for the Portuguese Language","abstract":"Computational medicine research requires clinical data for training and testing purposes, so the development of datasets composed of real hospital data is of utmost importance in this field. Most such data collections are in the English language, were collected in anglophone countries, and do not reflect other clinical realities, which increases the importance of such national datasets for projects that hope to positively impact public health. This paper presents a new Brazilian Clinical Dataset containing over 70,000 admissions from 10 hospitals in two Brazilian states, composed of a sum total of over 2.5 million free-text clinical notes alongside data pertaining to patient descriptors, prescription information, and exam results. This data was collected, organized, deidentified, and is being distributed via credentialed access for the use of the research community. In the course of presenting the new dataset, we explore the new dataset’s structure, population, and potential benefits for use in clinical AI tasks.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":1763370869,"compressed_storage_size":103110672,"resource_type":"Database","access_policy":"Contributor Review","topics":["prescriptions","exams","clinical notes","natural language processing","tertiary care"],"source_url":"https://physionet.org/content/brateca/1.1/"},{"public_project_uuid":"a53bff85-1aee-4872-9b68-42addd5fada2","slug":"simultaneous-measurements","version":"1.0.1","core_doi":"10.13026/ddzr-dw82","version_doi":"10.13026/zhns-t386","is_latest_version":false,"short_description":"Dataset to support comparison of usability and accuracy from simultaneous measurements collected from 13 subjects including five devices: NeXus-10 MKII, eMotion Faros 360°, Hexoskin Hx1, SOMNOTouch NIBP, Polar RS800 Multi.","publish_date":"2022-07-20","title":"Simultaneous physiological measurements with five devices at different cognitive and physical loads","abstract":"This open database contains physiological measurements from 13 adult and healthy subjects during a standardized experimental setup. The physiology was examined with five different devices (NeXus-10 MKII, eMotion Faros 360°, SOMNOtouch NIBP, Hexoskin Hx1, Polar RS800 Multi), which, in contrast to other studies, were recorded simultaneously. The experiment included a five minute baseline measurement, five minute walking on the treadmill, a five minute cognitive audio test, and a five minute uphill walking on the treadmill. Measurements included electrocardiography (ECG), photoplethysmography, accelerometry, oxygen saturation, respiration, heart rate, heart rate variability, and RR intervals sampled between 1 Hz and 8000 Hz. The dataset was originally generated during an investigation of functionality, accuracy, and usability of several ECG-measurement devices. This resource included the complete raw data files (EDF and HRM) and merged data with a manual expert annotation of heart beats and annotation files with information about the beginning of each experimental phase.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":1783802925,"compressed_storage_size":808909021,"resource_type":"Database","access_policy":"Open","topics":["photoplethysmogram","multiparameter","heart rate","holter","noise","ecg","temperature","movement","accelerometer","hrv","respiration"],"source_url":"https://physionet.org/content/simultaneous-measurements/1.0.1/"},{"public_project_uuid":"609c98f1-b406-4abe-b4cc-06c044e6d73a","slug":"gossis-1-eicu","version":"1.0.0","core_doi":"10.13026/drts-zb06","version_doi":"10.13026/gbmg-a531","is_latest_version":true,"short_description":"GOSSIS-1 is an in-hospital mortality prediction algorithm for critical care patients. GOSSIS-1 was trained using data from three countries. This dataset corresponds with the USA subset of the GOSSIS-1 dataset for the 2022 publication below.","publish_date":"2022-07-20","title":"GOSSIS-1-eICU, the eICU-CRD subset of the Global Open Source Severity of Illness Score (GOSSIS-1) dataset","abstract":"GOSSIS-1 is a modern, free, open-source in-hospital mortality prediction algorithm for critical care patients, achieving excellent discrimination and calibration across three countries (Australia, New Zealand and the USA). GOSSIS-1 was developed on two large datasets of critical care patients. This project contains the USA subset of patients derived from the eICU Collaborative Research Database (eICU-CRD). The dataset, which we call GOSSIS-1-eICU, consists of 131,051 unique patients from 204 hospitals from ICU admissions discharged in 2014-15. The code to create the dataset from eICU-CRD and generate GOSSIS-1 predictions are also available. This project contains: 1) the derived dataset from eICU-CRD, 2) the dataset with required missing data imputed and 3) the GOSSIS-1 in-hospital predictions (probabilities. The patientunitstayid and hospitalid eICU-CRD identifiers are included to allowing linking back to eICU-CRD. Training and test sets are identified to allow for direct comparisons of performance.
Existing question answering (QA) datasets derived from electronic health records (EHR) are artificially generated and consequently fail to capture realistic physician information needs. We present Discharge Summary Clinical Questions (DiSCQ), a newly curated question dataset composed of 2,000+ questions paired with the snippets of text (triggers) that prompted each question. The questions are generated by medical experts from 100+ MIMIC-III, version 1.4, discharge summaries. These discharge summaries overlap with the n2c2 challenge, so they are filled in with surrogate PHI. We analyze this dataset to characterize the types of information sought by medical experts. We also train baseline models for trigger detection and question generation (QG), paired with unsupervised answer retrieval over EHRs. Our baseline model is able to generate high quality questions in over 62% of cases when prompted with human selected triggers. We release this dataset (and a link to all code to reproduce baseline model results) to facilitate further research into realistic clinical QA and QG.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":792157,"compressed_storage_size":189834,"resource_type":"Database","access_policy":"Credentialed","topics":["machine learning","question answering","question generation"],"source_url":"https://physionet.org/content/discq/1.0/"},{"public_project_uuid":"419909be-8f89-45e0-8f66-c2ad2d4aff1d","slug":"challenge-2020","version":"1.0.2","core_doi":"10.13026/m77n-sx13","version_doi":"10.13026/dvyd-kd57","is_latest_version":true,"short_description":"The goal of the 2020 PhysioNet - Computing in Cardiology Challenge is to design and implement a working, open-source algorithm that can automatically identify cardiac abnormalities in 12-lead ECG recordings.","publish_date":"2022-07-29","title":"Classification of 12-lead ECGs: The PhysioNet/Computing in Cardiology Challenge 2020","abstract":"The electrocardiogram (ECG) is a non-invasive representation of the electrical activity of the heart from electrodes placed on the surface of the torso. The standard 12-lead ECG has been widely used to diagnose a variety of cardiac abnormalities such as cardiac arrhythmias, and predicts cardiovascular morbidity and mortality [1]. The early and correct diagnosis of cardiac abnormalities can increase the chances of successful treatments [2]. However, manual interpretation of the electrocardiogram is time-consuming, and requires skilled personnel with a high degree of training [3].
\n\nAutomatic detection and classification of cardiac abnormalities can assist physicians in the diagnosis of the growing number of ECGs recorded. Over the last decade, there have been increasing numbers of attempts to stimulate 12-lead ECG classification. Many of these algorithms seem to have the potential for accurate identification of cardiac abnormalities. However, most of these methods have only been tested or developed in single, small, or relatively homogeneous datasets. The PhysioNet/Computing in Cardiology Challenge 2020 provides an opportunity to address this problem by providing data from a wide set of sources.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":8100183042,"compressed_storage_size":4814544948,"resource_type":"Challenge","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/challenge-2020/1.0.2/"},{"public_project_uuid":"415011eb-5c0e-4b76-b237-fa0350b63f52","slug":"challenge-2021","version":"1.0.3","core_doi":"10.13026/gt86-a263","version_doi":"10.13026/34va-7q14","is_latest_version":true,"short_description":"Will Two Do? Varying Dimensions in Electrocardiography: the PhysioNet/Computing in Cardiology Challenge 2021","publish_date":"2022-07-29","title":"Will Two Do? Varying Dimensions in Electrocardiography: The PhysioNet/Computing in Cardiology Challenge 2021","abstract":"The electrocardiogram (ECG) is a non-invasive representation of the electrical activity of the heart. Although the twelve-lead ECG is the standard diagnostic screening system for many cardiological issues, the limited accessibility of twelve-lead ECG devices provides a rationale for smaller, lower-cost, and easier to use devices. While single-lead ECGs are limiting [1], reduced-lead ECG systems hold promise, with evidence that subsets of the standard twelve leads can capture useful information [2], [3], [4] and even be comparable to twelve-lead ECGs in some limited contexts. In 2017 we challenged the public to classify AF from a single-lead ECG, and in 2020 we challenged the public to diagnose a much larger number of cardiac problems using twelve-lead recordings. However, there is limited evidence to demonstrate the utility of reduced-lead ECGs for capturing a wide range of diagnostic information.
\n\nIn this year’s Challenge, we ask the following question: ‘Will two do?’ This year’s Challenge builds on last year’s Challenge [5], which asked participants to classify cardiac abnormalities from twelve-lead ECGs. We are asking you to build an algorithm that can classify cardiac abnormalities from twelve-lead, six-lead, four-lead, three-lead, and two-lead ECGs. We will test each algorithm on databases of these reduced-lead ECGs, and the differences in performances of the algorithms on these databases will reveal the utility of reduced-lead ECGs in comparison to standard twelve-lead EGCs.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":13575090358,"compressed_storage_size":0,"resource_type":"Challenge","access_policy":"Open","topics":["cardiac abnormalities","challenge","multilead ecgs","classification","competition"],"source_url":"https://physionet.org/content/challenge-2021/1.0.3/"},{"public_project_uuid":"d1f29bcd-d994-4395-84b4-29d4d67b0170","slug":"cded","version":"1.0.1","core_doi":"10.13026/bpnv-4b88","version_doi":"10.13026/00bm-0x81","is_latest_version":true,"short_description":"Type 2 diabetes increases risk for cerebrovascular disease, cognitive and mobility decline in older people. This project evaluated relationship between diabetes, inflammation cerebrovascular reactivity and functional outcomes.","publish_date":"2022-08-05","title":"Cerebromicrovascular Disease in Elderly with Diabetes","abstract":"This dataset was collected as part of a prospective observational study to evaluate the effects of type 2 diabetes mellitus (DM) on cerebral vasoregulation, perfusion and functional outcomes, measured by blood flow responses to hypocapnia and hypercapnia, Valsalva maneuver, head-up tilt, and sit-to-stand test. The dataset comprises of observations from 69 diabetic and control participants (aged 55 to 75 years) with continuous measurements of cerebral blood flow using transcranial Doppler and MRI (magnetic resonance imaging), heart rate, blood pressure, and respiratory parameters, balance, walking, laboratory and retinopathy measures at baseline, and 41 subjects who completed the two-years of follow-up. Regional gray, white matter and cerebrospinal fluid volumes were quantified using a segmentation method applied on T1- and T2- weighted images and perfusion maps, using a continuous arterial spin labeling (CASL) images at 3 Tesla MRI. White matter integrity was determined from fluid attenuated inversion recovery (FLAIR) and diffusion tensor imaging (DTI) MRI. Dynamics of cerebral vasoregulation to CO2 challenge and orthostatic stress were measured using Transcranial Doppler ultrasound (TCD).
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":3941743964,"compressed_storage_size":2211624535,"resource_type":"Database","access_policy":"Open","topics":["gait and balance","brain perfusion","inflammation","type 2 diabetes","cerebromiscovascular disease","cerebral vasoreactivity","brain atrophy","perfusion mri"],"source_url":"https://physionet.org/content/cded/1.0.1/"},{"public_project_uuid":"ecaf3ef4-9ac2-454e-a438-1468a37b6c5e","slug":"cerebral-perfusion-diabetes","version":"1.0.1","core_doi":"10.13026/pgn6-6k53","version_doi":"10.13026/whjz-e968","is_latest_version":true,"short_description":"Dataset collected during a study on type 2 diabetes on brain blood flow, vasoreactivity and functional outcomes (gait and balance) using TCD, MRI perfusion and foot pressure distribution and gait measures.","publish_date":"2022-08-05","title":"Cerebral perfusion and cognitive decline in type 2 diabetes","abstract":"This dataset was collected as part of a study to explore vasoregulation and blood flow in patients with type 2 Diabetes mellitus. It comprises cerebral blood, autonomic functions, laboratory and gait variables in 70 patients with type 2 DM and 70 healthy controls (50-85 years old). Diabetes mellitus is a prevalent, costly condition with high morbidity and mortality that affects nearly 6.3% of the population in the United States. Diabetes mellitus alters the permeability of the blood-brain barrier, thus affecting regional metabolism and microcirculatory regulation. Predictors of cerebrovascular complications of diabetes that are evidence-based on cerebral blood flow, MRI imaging, and cognitive testing are lacking.
\n\nBlood flow velocities were measured in the anterior and middle cerebral arteries using transcranial Doppler ultrasound. Cerebral vasoregulation was evaluated by comparing blood flow velocity responses during hypocapnia and hypercapnia, Valsalva maneuver, head-up tilt and sit-to-stand test using simultaneous recordings of cardiovascular variables, blood flow velocity in the anterior and middle cerebral artery using transcranial Doppler and respiratory variables. We applied continuous arterial spin labeling at 3 Tesla MRI to evaluate regional distribution of cerebral blood flow and vasomotor reactivity to CO2. Cognitive and executive functions were assessed via a battery of psychometric measures. Gait and balance were characterized from dynamics of foot pressure distribution and center of pressure displacement.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":13990755868,"compressed_storage_size":5435731605,"resource_type":"Database","access_policy":"Open","topics":["vasoregulation","brain","diabetes"],"source_url":"https://physionet.org/content/cerebral-perfusion-diabetes/1.0.1/"},{"public_project_uuid":"1f60f35a-671a-4833-975d-fb884df09966","slug":"wfdb-python","version":"4.0.0","core_doi":"10.13026/jt5e-gq15","version_doi":"10.13026/mmpm-2v55","is_latest_version":false,"short_description":"Tools for working with waveforms in Python.","publish_date":"2022-08-09","title":"Waveform Database Software Package (WFDB) for Python","abstract":"Physiological waveforms - such as electrocardiograms (ECG), electroencephalograms (EEG), electromyograms (EMG) - are generated during the course of routine care. These signals contain information that can be used to understand underlying conditions of health. Effective processing and analysis of physiological data requires specialized software. The waveform-database (WFDB) package for Python is a library of tools for reading, writing, and processing physiological signals and annotations, written in the Python programming language. Core components of the Python package are loosely based on the specifications of the original C-language WFDB software. We aim to implement as many of the core WFDB features as possible, with user-friendly APIs. Additional features are added over time.
","license":{"name":"MIT License"},"dua":null,"main_storage_size":245232497,"compressed_storage_size":104202684,"resource_type":"Software","access_policy":"Open","topics":["wfdb","python","waveform"],"source_url":"https://physionet.org/content/wfdb-python/4.0.0/"},{"public_project_uuid":"c5631663-96a7-40d2-b8ab-8d9df308e45f","slug":"haim-multimodal","version":"1.0.0","core_doi":"10.13026/dxcx-n572","version_doi":"10.13026/e5hj-1229","is_latest_version":false,"short_description":"Code for generating the HAIM multimodal dataset of MIMIC-IV clinical data and x-rays","publish_date":"2022-08-17","title":"Code for generating the HAIM multimodal dataset of MIMIC-IV clinical data and x-rays","abstract":"A multimodal combination of the MIMIC-IV v1.0.0 and MIMIC Chest X-ray (MIMIC-CXR-JPG) v2.0.0 databases filtered to only include patients that have at least one chest X-ray performed with the goal of validating multi-modal predictive analytics in healthcare operations can be generated with the present resource. This multimodal dataset generated through this code contains 34,540 individual patient files in the form of "pickle" Python object structures, which covers a total of 7,279 hospitalization stays involving 6,485 unique patients. Additionally, code to extract feature embeddings as well as the list of pre-processed features are included in this repository.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":14901743550,"compressed_storage_size":5513581074,"resource_type":"Software","access_policy":"Credentialed","topics":["multimodality","database","code"],"source_url":"https://physionet.org/content/haim-multimodal/1.0.0/"},{"public_project_uuid":"4c293910-5a17-45a5-8b3c-9435167658e8","slug":"lung-segment-mimic-cxr","version":"1.0.0","core_doi":"10.13026/k0v0-6j70","version_doi":"10.13026/d9zv-4d92","is_latest_version":true,"short_description":"A chest x-rays segmentation dataset derived from MIMIC-CXR based on deep learning algorithm and human examination.","publish_date":"2022-08-18","title":"Chest X-ray segmentation images based on MIMIC-CXR","abstract":"As more and more artificial intelligence (AI) or deep learning technologies have been applied to medical image applications such as radiological finding identification in chest X-rays (CXRs), the interpretability of the prediction model is crucial for building trust in AI. In pulmonary pathology detection, the CXR images with proper anatomical segmentations could aid in interpreting the models. However, the accuracy of the auto-segmentation algorithms was not high enough to create such a benchmark. In this project, we provided segmentation results of 1,141 frontal-view CXRs randomly selected from the MIMIC-CXR database. These CXRs were first processed into a pair of segmented images with the lung lobes and the rest parts by deep learning-based algorithms. We then manually filtered out the incorrect segmentation results. The segmented images maybe helpful for model interpretability.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":35981639,"compressed_storage_size":33290735,"resource_type":"Database","access_policy":"Credentialed","topics":["cxr","segmentation","chest x-rays"],"source_url":"https://physionet.org/content/lung-segment-mimic-cxr/1.0.0/"},{"public_project_uuid":"58954686-2ffb-4570-a060-9d1c847ba949","slug":"ptb-xl","version":"1.0.2","core_doi":"10.13026/6sec-a640","version_doi":"10.13026/zx4k-te85","is_latest_version":false,"short_description":"The PTB-XL ECG dataset is a large dataset of 21801 clinical 12-lead ECGs from 18869 patients of 10 second length. The raw signal data has been annotated by up to two cardiologists with 71 different ECG statements and is supplemented by rich metadata.","publish_date":"2022-08-18","title":"PTB-XL, a large publicly available electrocardiography dataset","abstract":"Electrocardiography (ECG) is a key diagnostic tool to assess the cardiac condition of a patient. Automatic ECG interpretation algorithms as diagnosis support systems promise large reliefs for the medical personnel - only on the basis of the number of ECGs that are routinely taken. However, the development of such algorithms requires large training datasets and clear benchmark procedures. In our opinion, both aspects are not covered satisfactorily by existing freely accessible ECG datasets.
\n\nThe PTB-XL ECG dataset is a large dataset of 21801 clinical 12-lead ECGs from 18869 patients of 10 second length. The raw waveform data was annotated by up to two cardiologists, who assigned potentially multiple ECG statements to each record. The in total 71 different ECG statements conform to the SCP-ECG standard and cover diagnostic, form, and rhythm statements. To ensure comparability of machine learning algorithms trained on the dataset, we provide recommended splits into training and test sets. In combination with the extensive annotation, this turns the dataset into a rich resource for the training and the evaluation of automatic ECG interpretation algorithms. The dataset is complemented by extensive metadata on demographics, infarction characteristics, likelihoods for diagnostic ECG statements as well as annotated signal properties.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":3184832604,"compressed_storage_size":1840324760,"resource_type":"Database","access_policy":"Open","topics":["ptb-xl","ecg","ptb","electrocardiography"],"source_url":"https://physionet.org/content/ptb-xl/1.0.2/"},{"public_project_uuid":"a678cdca-3df0-4151-898d-a81cd0c0600c","slug":"haim-multimodal","version":"1.0.1","core_doi":"10.13026/dxcx-n572","version_doi":"10.13026/3f8d-qe93","is_latest_version":true,"short_description":"Code for generating the HAIM multimodal dataset of MIMIC-IV clinical data and x-rays","publish_date":"2022-08-23","title":"Code for generating the HAIM multimodal dataset of MIMIC-IV clinical data and x-rays","abstract":"A multimodal combination of the MIMIC-IV v1.0.0 and MIMIC Chest X-ray (MIMIC-CXR-JPG) v2.0.0 databases filtered to only include patients that have at least one chest X-ray performed with the goal of validating multi-modal predictive analytics in healthcare operations can be generated with the present resource. This multimodal dataset generated through this code contains 34,540 individual patient files in the form of "pickle" Python object structures, which covers a total of 7,279 hospitalization stays involving 6,485 unique patients. Additionally, code to extract feature embeddings as well as the list of pre-processed features are included in this repository.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":14901743550,"compressed_storage_size":5513581074,"resource_type":"Software","access_policy":"Credentialed","topics":["code","database","multimodality"],"source_url":"https://physionet.org/content/haim-multimodal/1.0.1/"},{"public_project_uuid":"5e367dbf-79d0-4f54-b67a-5bde9120a2c2","slug":"ecg-arrhythmia","version":"1.0.0","core_doi":"10.13026/92ks-sq55","version_doi":"10.13026/wgex-er52","is_latest_version":true,"short_description":"A 12-lead electrocardiogram database for arrhythmia research covering more than 10,000 patients","publish_date":"2022-08-24","title":"A large scale 12-lead electrocardiogram database for arrhythmia study","abstract":"This newly inaugurated research database for 12-lead electrocardiogram (ECG) signals was created under the auspices of Chapman University, Shaoxing People’s Hospital (Shaoxing Hospital Zhejiang University School of Medicine), and Ningbo First Hospital. It aims to enable the scientific community in conducting new studies on arrhythmia and other cardiovascular conditions. Certain types of arrhythmias, such as atrial fibrillation, have a pronounced negative impact on public health, quality of life, and medical expenditures. As a non-invasive test, ECG is a major and vital diagnostic tool for detecting these conditions. This practice, however, generates large amounts of data, the analysis of which requires considerable time and effort by human experts. Modern machine learning and statistical tools can be trained on high quality, large data to achieve exceptional levels of automated diagnostic accuracy. Thus, we collected and disseminated this novel database that contains 12-lead ECGs of 45,152 patients with a 500 Hz sampling rate that features multiple common rhythms and additional cardiovascular conditions, all labeled by professional experts. The dataset can be used to design, compare, and fine-tune new and classical statistical and machine learning techniques in studies focused on arrhythmia and other cardiovascular conditions.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":5458944594,"compressed_storage_size":2498709074,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/ecg-arrhythmia/1.0.0/"},{"public_project_uuid":"4fadc2ca-6f2e-44e6-b956-5921a0696136","slug":"virtual-reality-piloting","version":"1.0.0","core_doi":"10.13026/tjpc-fm02","version_doi":"10.13026/azwa-ge48","is_latest_version":true,"short_description":"Physiologic and flight performance data from a virtual reality task. Data for the CogPilot Data Challenge.","publish_date":"2022-08-25","title":"Multimodal Physiological Monitoring During Virtual Reality Piloting Tasks","abstract":"This dataset includes multimodal physiologic, flight performance, and user interaction data streams, collected as participants performed virtual flight tasks of varying difficulty. In virtual reality, individuals flew an "Instrument Landing System" (ILS) protocol, in which they had to land an aircraft mostly relying on the cockpit instrument readings. Participants were presented with four levels of difficulty, which were generated by varying wind speed, turbulence, and visibility. Each of the participants performed 12 runs, split into 3 blocks of four consecutive runs, one run at each difficulty, in a single experimental session. The sequence of difficulty levels was presented in a counterbalanced manner across blocks. Flight performance was quantified as a function of horizontal and vertical deviation from an ideal path towards the runway as well as deviation from the prescribed ideal speed of 115 knots. Multimodal physiological signals were aggregated and synchronized using Lab Streaming Layer. Descriptions of data quality are provided to assess each data stream. The starter code provides examples of loading and plotting the time synchronized data streams, extracting sample features from the eye tracking data, and building models to predict pilot performance from the physiology data streams.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":47629407001,"compressed_storage_size":8951131229,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/virtual-reality-piloting/1.0.0/"},{"public_project_uuid":"6725c9d9-b9d7-4ec0-88dd-cfdf99ffab71","slug":"nosocomialriskdata","version":"1.0","core_doi":"10.13026/p7ww-c390","version_doi":"10.13026/pm60-0g49","is_latest_version":true,"short_description":"Text-based Longitudinal Data for Predicting Nosocomial Disease Risk as used by CANTRIP.","publish_date":"2022-09-15","title":"Nosocomial Risk Datasets from MIMIC-III","abstract":"Reliable longitudinal risk prediction for hospitalized patients is needed to provide quality care. Our goal is to foster the development of generalizable models capable of leveraging clinical notes to predict healthcare-associated diseases 24–96 hours in advance. We developed data to explore the problem of predicting the risk of hospital acquired (occurring 48 hours after admission) acute kidney injury, pressure injury, or anemia 24 hours before it is implicated by the patient’s chart, labs, or notes. We relied on the MIMIC-III critical care database and extract distinct positive and negative cohorts for each disease. We retrospectively determine the date-of-event using structured and unstructured criteria so that it may be used as a form of indirect supervision to train and evaluate automatic systems for predicting disease risk using clinical notes. This data was used as the experimental basis for the CANTRIP project.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1590205887,"compressed_storage_size":324022675,"resource_type":"Database","access_policy":"Credentialed","topics":["pressure injury","natural language processing","risk prediction","acute kidney injury","deep learning","anemia","forecasting"],"source_url":"https://physionet.org/content/nosocomialriskdata/1.0/"},{"public_project_uuid":"eb8d6e34-499c-48b0-b1d1-490c531a4fb2","slug":"fetalheartsounddata","version":"1.0","core_doi":"10.13026/at0m-2g90","version_doi":"10.13026/9vvw-cx05","is_latest_version":true,"short_description":"The IIScFHSDB has 60 fetal phonocardiography recordings obtained in a hospital setting with the objective to provide fPCG recordings with clinical noise settings for development of signal processing algorithms for FHR determination and denoising fPCG","publish_date":"2022-09-15","title":"Indian Institute of Science Fetal Heart Sound Database (IIScFHSDB)","abstract":"The Indian Institute of Science Fetal Heart Sound Database (IIScFHSDB) has 60 fetal phonocardiography (fPCG) recordings obtained from expecting mothers of age 18-27 years of age with gestation weeks 30-40. The recordings are made at St. John’s Hospital, Bangalore with an electronic stethoscope. The data is collected from the antenatal and labor wards by placing the electronic stethoscope on the lower abdomen of the subject. The fPCG recordings are obtained from 60 subjects with the gain of 500X and the average duration of the recordings is 8 minutes with a sampling frequency of 2 kHz.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":395542600,"compressed_storage_size":134498771,"resource_type":"Database","access_policy":"Open","topics":["1 fetal heart sound 2 fhr"],"source_url":"https://physionet.org/content/fetalheartsounddata/1.0/"},{"public_project_uuid":"04f3cd90-04d4-4a18-affa-4421d57a7d5d","slug":"big-ideas-glycemic-wearable","version":"1.0.0","core_doi":"10.13026/w591-tp72","version_doi":"10.13026/d742-0j82","is_latest_version":false,"short_description":"Glucose measurements and wrist-worn wearable sensor data from highnormoglycemic participants.","publish_date":"2022-09-15","title":"BIG IDEAs Lab Glycemic Variability and Wearable Device Data","abstract":"This study aimed to determine the feasibility and effectiveness of wearable devices in detecting early physiological changes prior to the development of prediabetes [1-3]. The study generated digital biomarkers for remote, mHealth-based prediabetes and hyperglycemia risk to classify which individuals should undergo further clinical testing. The primary inclusion criteria were subjects aged 35-65 years, inclusive, including only post-menopausal females, with a point of care A1C measurement between 5.2-6.4%, inclusive. Blood was collected during the study for measurement of glucose, hemoglobin A1C, lipoproteins, and triglycerides. Participants wore a Dexcom 6 continuous glucose monitor (CGM) and an Empatica E4 wristband for 10 days while receiving a standardized breakfast meal every other day. At the end of the 10 days, the participant returned to the clinic for an oral glucose tolerance test (OGTT). Research data collected includes physiological measurements from wearable devices such as heart rate, accelerometry, and electrodermal conductance.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":36649468680,"compressed_storage_size":5016557349,"resource_type":"Database","access_policy":"Open","topics":["biomedical engineering","pre-diabetes","biomarkers"],"source_url":"https://physionet.org/content/big-ideas-glycemic-wearable/1.0.0/"},{"public_project_uuid":"31c105a5-6b2a-4b7e-a51a-00e2b91fbea4","slug":"grabmyo","version":"1.0.1","core_doi":"10.13026/cadk-kw71","version_doi":"10.13026/701k-gs64","is_latest_version":false,"short_description":"Open-access dataset of electromyogram (EMG) recordings collected from the wrist and forearm muscles of 43 people while they performed hand gestures.","publish_date":"2022-09-21","title":"Gesture Recognition and Biometrics ElectroMyogram (GRABMyo)","abstract":"We present the Gesture Recognition and Biometrics ElectroMyogram (GRABMyo) dataset, an open-access dataset of electromyogram (EMG) recordings collected from the wrist and forearm muscles while performing hand gestures. Data were collected from 43 healthy participants (age range: 24-35 years) on three different days while performing 16 hand and finger gestures in identical experimental sessions on each day. The GRABMyo dataset can be used for research on: 1) EMG-based biometrics for personal identification and verification, and 2) EMG-based gesture recognition for neurorehabilitation and home applications. The large sample size will provide sufficient power for establishing results, specifically for applications such as biometrics. Further, it will be useful for subject-independent applications such as generalized classification models for gesture recognition. The multiday recordings will provide results that will be consistent over a longer duration, crucial for the reliability of EMG-based wearable devices.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":10115496087,"compressed_storage_size":9820453229,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/grabmyo/1.0.1/"},{"public_project_uuid":"1b258ce8-2a7c-4b2f-a054-493effda4540","slug":"mimic3-carevue","version":"1.4","core_doi":"10.13026/yj38-pn40","version_doi":"10.13026/8a4q-w170","is_latest_version":true,"short_description":"A subset of the MIMIC-III Clinical Database containing only patients admitted from 2001 - 2008.","publish_date":"2022-09-21","title":"MIMIC-III Clinical Database CareVue subset","abstract":"MIMIC-III is a database of critically ill patients admitted to an intensive care unit (ICU) at the Beth Israel Deaconess Medical Center (BIDMC) in Boston, MA. MIMIC-III has seen broad use, and was updated with the release of MIMIC-IV. MIMIC-IV contains more contemporaneous stays, higher granularity data, and expanded domains of information. To maximize the sample size of MIMIC-IV, the database overlaps with MIMIC-III, and specifically both databases contain the same admissions which occurred between 2008 - 2012. This overlap complicates analyses of the two databases simultaneously. Here we provide a subset of MIMIC-III containing patients who are not in MIMIC-IV. The goal of this project is to simplify the combination of MIMIC-III with MIMIC-IV.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":2571411520,"compressed_storage_size":2540260544,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimic3-carevue/1.4/"},{"public_project_uuid":"a6c35622-fdf0-4826-b34f-ad2e3da43d6a","slug":"vitaldb","version":"1.0.0","core_doi":"10.13026/w758-nw21","version_doi":"10.13026/czw8-9p62","is_latest_version":true,"short_description":"VitalDB, a high-fidelity multi-parameter vital signs database in surgical patients","publish_date":"2022-09-21","title":"VitalDB, a high-fidelity multi-parameter vital signs database in surgical patients","abstract":"In modern anesthesia, multiple medical devices are used simultaneously to comprehensively monitor real-time vital signs to optimize patient care and improve surgical outcomes. However, interpreting the dynamic changes of time-series biosignals and their correlations is a difficult task even for experienced anesthesiologists. Recent advanced machine learning technologies have shown promising results in biosignal analysis, however, research and development in this area is relatively slow due to the lack of biosignal datasets for machine learning. The VitalDB (Vital Signs DataBase) is an open dataset created specifically to facilitate machine learning studies related to monitoring vital signs in surgical patients. This dataset contains high-resolution multi-parameter data from 6,388 cases, including 486,451 waveform and numeric data tracks of 196 intraoperative monitoring parameters, 73 perioperative clinical parameters, and 34 time-series laboratory result parameters. All data is stored in the public cloud after anonymization. The dataset can be freely accessed and analysed using application programming interfaces and Python library. The VitalDB public dataset is expected to be a valuable resource for biosignal research and development.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":102456727132,"compressed_storage_size":101136300876,"resource_type":"Database","access_policy":"Open","topics":["vitaldb","waveform","intraoperative","ecg","anesthesia","biosignal"],"source_url":"https://physionet.org/content/vitaldb/1.0.0/"},{"public_project_uuid":"e1dd76d0-76a9-4882-9b87-74db578ca81e","slug":"task-1-3-soap-note-tag","version":"1.0.0","core_doi":"10.13026/5qe0-x984","version_doi":"10.13026/wks0-w041","is_latest_version":true,"short_description":"We introduce a hierarchical annotation suite of tasks addressing clinical text understanding, reasoning and abstraction over evidence, and diagnosis summarization. One task is section tagging major section and the other task is diagnosis generation.","publish_date":"2022-09-30","title":"Tasks 1 and 3 from Progress Note Understanding Suite of Tasks: SOAP Note Tagging and Problem List Summarization","abstract":"Applying methods in natural language processing on electronic health records (EHR) data is a growing field. Existing corpus and annotation focus on modelling textual features and relation prediction [1] . However, there is a paucity of annotated corpus built to model clinical diagnostic reasoning, a process that involves text understanding, domain knowledge abstraction and reasoning, and clinical text generation. The datasets here support a hierarchical annotation schema with two out of the three stages available to address clinical text understanding and text generation. The datasets provided here are for individual tasks in Stages 1 and 3. The task for Stage 2 was previously accepted as part of the National NLP Clinical Challenges (n2c2) and may be retrieved from the n2c2 challenge website.
\n\nThe annotated corpus is based on an extensive collection of Intensive Care Unit progress notes, a type of EHR documentation that is collected in time series in a problem-oriented format. The progress notes were sourced from MIMIC-III. The conventional format for a progress note follows a Subjective, Objective, Assessment and Plan heading (SOAP). The novel suite of tasks was designed to train and evaluate future NLP models for clinical text understanding, clinical knowledge representation, inference, and summarization. The ultimate goal of these datasets is to advance the development and evaluation of NLP models for clinical applications that lead to AI-assisted clinical decision support and reduce medical errors.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":6685766,"compressed_storage_size":2002807,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/task-1-3-soap-note-tag/1.0.0/"},{"public_project_uuid":"8f20c35a-c98b-4736-949f-08dc6b1f6c69","slug":"kinecal","version":"1.0.1","core_doi":"10.13026/8bsb-ch59","version_doi":"10.13026/z037-6191","is_latest_version":false,"short_description":"A dataset for balance falls-risk assessment and balance impairment analysis","publish_date":"2022-10-31","title":"KINECAL","abstract":"The field of human action recognition has made great strides in recent years, much helped by the availability of a wide variety of datasets that use Kinect to record human movement. Conversely, progress towards using Kinect in clinical practice has been hampered by the lack of appropriate data. In particular, datasets that contain clinically significant movements and appropriate metadata. This paper proposes a dataset to address this issue, namely KINECAL. It contains the recordings of 90 individuals carrying out 11 movements, commonly used in the clinical assessment of balance. The dataset contains relevant metadata, including clinical labelling, falls history and postural sway metrics. In addition, the recordings were made in both the lab and in informal settings.
","license":{"name":"Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":192423495337,"compressed_storage_size":92063952913,"resource_type":"Database","access_policy":"Open","topics":["age-related changes","falls-risk","postural sway","posturography","balance","clinical tests"],"source_url":"https://physionet.org/content/kinecal/1.0.1/"},{"public_project_uuid":"e6c5a8bb-cad5-44a7-a4a7-ad10f866b653","slug":"globem","version":"1.0","core_doi":"10.13026/889w-p048","version_doi":"10.13026/jvtb-2d81","is_latest_version":false,"short_description":"GLOBEM datasets contain the first released multi-year mobile and wearable sensing datasets from 2018 to 2021, containing 705 person-years and 497 unique participants.","publish_date":"2022-11-04","title":"GLOBEM Dataset: Multi-Year Datasets for Longitudinal Human Behavior Modeling Generalization","abstract":"We present the first multi-year mobile sensing datasets. Our multi-year data collection studies span four years (10 weeks each year, from 2018 to 2021). The four datasets contain data collected from 705 person-years (497 unique participants) with diverse racial, ability, and immigrant backgrounds. Each year, participants would install a mobile app on their phones and wear a fitness tracker. The app and wearable device passively track multiple sensor streams in the background 24×7, including location, phone usage, calls, Bluetooth, physical activity, and sleep behavior. In addition, participants completed weekly short surveys and two comprehensive surveys on health behaviors and symptoms, social well-being, emotional states, mental health, and other metrics. Our dataset analysis indicates that our datasets capture a wide range of daily human routines, and reveal insights between daily behaviors and important well-being metrics (e.g., depression status). We envision our multi-year datasets can support the ML community in developing generalizable longitudinal behavior modeling algorithms.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":3127032872,"compressed_storage_size":877628504,"resource_type":"Database","access_policy":"Credentialed","topics":["ubiquitous computing","passive mobile sensing","human behavior modeling","well-being","health"],"source_url":"https://physionet.org/content/globem/1.0/"},{"public_project_uuid":"c63b8ff1-6ecd-4548-8b7b-b6e012d72fe3","slug":"ptb-xl","version":"1.0.3","core_doi":"10.13026/6sec-a640","version_doi":"10.13026/kfzx-aw45","is_latest_version":true,"short_description":"The PTB-XL ECG dataset is a large dataset of 21801 clinical 12-lead ECGs from 18869 patients of 10 second length. The raw signal data has been annotated by up to two cardiologists with 71 different ECG statements and is supplemented by rich metadata.","publish_date":"2022-11-09","title":"PTB-XL, a large publicly available electrocardiography dataset","abstract":"Electrocardiography (ECG) is a key diagnostic tool to assess the cardiac condition of a patient. Automatic ECG interpretation algorithms as diagnosis support systems promise large reliefs for the medical personnel - only on the basis of the number of ECGs that are routinely taken. However, the development of such algorithms requires large training datasets and clear benchmark procedures. In our opinion, both aspects are not covered satisfactorily by existing freely accessible ECG datasets.
\n\nThe PTB-XL ECG dataset is a large dataset of 21799 clinical 12-lead ECGs from 18869 patients of 10 second length. The raw waveform data was annotated by up to two cardiologists, who assigned potentially multiple ECG statements to each record. The in total 71 different ECG statements conform to the SCP-ECG standard and cover diagnostic, form, and rhythm statements. To ensure comparability of machine learning algorithms trained on the dataset, we provide recommended splits into training and test sets. In combination with the extensive annotation, this turns the dataset into a rich resource for the training and the evaluation of automatic ECG interpretation algorithms. The dataset is complemented by extensive metadata on demographics, infarction characteristics, likelihoods for diagnostic ECG statements as well as annotated signal properties.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":3181524022,"compressed_storage_size":1839504686,"resource_type":"Database","access_policy":"Open","topics":["ecg","ptb-xl","ptb","electrocardiography"],"source_url":"https://physionet.org/content/ptb-xl/1.0.3/"},{"public_project_uuid":"6ea0e320-4a01-4a30-8992-c7223efd4194","slug":"mimiciv","version":"2.1","core_doi":"10.13026/07hj-2a80","version_doi":"10.13026/rrgf-xw32","is_latest_version":false,"short_description":"Database of patients admitted to the Beth Israel Deaconess Medical Center","publish_date":"2022-11-16","title":"MIMIC-IV","abstract":"Retrospectively collected medical data has the opportunity to improve patient care through knowledge discovery and algorithm development. Broad reuse of medical data is desirable for the greatest public good, but data sharing must be done in a manner which protects patient privacy. The Medical Information Mart for Intensive Care (MIMIC)-III database provided critical care data for over 40,000 patients admitted to intensive care units at the Beth Israel Deaconess Medical Center (BIDMC). Importantly, MIMIC-III was deidentified, and patient identifiers were removed according to the Health Insurance Portability and Accountability Act (HIPAA) Safe Harbor provision. MIMIC-III has been integral in driving large amounts of research in clinical informatics, epidemiology, and machine learning. Here we present MIMIC-IV, an update to MIMIC-III, which incorporates contemporary data and improves on numerous aspects of MIMIC-III. MIMIC-IV adopts a modular approach to data organization, highlighting data provenance and facilitating both individual and combined use of disparate data sources. MIMIC-IV is intended to carry on the success of MIMIC-III and support a broad set of applications within healthcare.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":7533446522,"compressed_storage_size":7455618772,"resource_type":"Database","access_policy":"Credentialed","topics":["critical care","machine learning","mimic","intensive care unit"],"source_url":"https://physionet.org/content/mimiciv/2.1/"},{"public_project_uuid":"21b37a3d-5d60-4a2c-bdab-a34fcb51a095","slug":"cxr-pro","version":"1.0.0","core_doi":"10.13026/n5sr-av47","version_doi":"10.13026/frag-yn96","is_latest_version":true,"short_description":"CXR-PRO is an adaptation of the MIMIC-CXR dataset (consisting of chest radiographs and their associated free-text radiology reports) with references to non-existent priors removed.","publish_date":"2022-11-23","title":"CXR-PRO: MIMIC-CXR with Prior References Omitted","abstract":"CXR-PRO is an adaptation of the MIMIC-CXR dataset that omits references to prior radiology reports. Consisting of 374,139 free-text radiology reports and associated chest radiographs, CXR-PRO addresses the issue of hallucinated references to priors produced by radiology report generation models. By removing nearly all prior references in MIMIC-CXR, CXR-PRO, when used as training data for report generation models, is capable of broadly improving the factual consistency and accuracy of generated reports. More generally, this dataset aims to support a wide body of research in medical image analysis and natural language processing. MIMIC-CXR is a de-identified dataset, so no protected health information (PHI) is included.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1575798327,"compressed_storage_size":303595070,"resource_type":"Database","access_policy":"Credentialed","topics":["generation","large language models","free-text radiology reports","references to priors","retrieval"],"source_url":"https://physionet.org/content/cxr-pro/1.0.0/"},{"public_project_uuid":"09e18e74-2d17-45c0-b529-be652748f5e5","slug":"grabmyo","version":"1.0.2","core_doi":"10.13026/cadk-kw71","version_doi":"10.13026/dzhr-nj10","is_latest_version":false,"short_description":"Open-access dataset of electromyogram (EMG) recordings collected from the wrist and forearm muscles of 43 people while they performed hand gestures.","publish_date":"2022-12-02","title":"Gesture Recognition and Biometrics ElectroMyogram (GRABMyo)","abstract":"We present the Gesture Recognition and Biometrics ElectroMyogram (GRABMyo) dataset, an open-access dataset of electromyogram (EMG) recordings collected from the wrist and forearm muscles while performing hand gestures. Data were collected from 43 healthy participants (age range: 24-35 years) on three different days while performing 16 hand and finger gestures in identical experimental sessions on each day. The GRABMyo dataset can be used for research on: 1) EMG-based biometrics for personal identification and verification, and 2) EMG-based gesture recognition for neurorehabilitation and home applications. The large sample size will provide sufficient power for establishing results, specifically for applications such as biometrics. Further, it will be useful for subject-independent applications such as generalized classification models for gesture recognition. The multiday recordings will provide results that will be consistent over a longer duration, crucial for the reliability of EMG-based wearable devices.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":10115485114,"compressed_storage_size":9820449877,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/grabmyo/1.0.2/"},{"public_project_uuid":"3ca3f256-fad6-4b86-a446-dd3c053d7842","slug":"radqa","version":"1.0.0","core_doi":"10.13026/nz91-gb70","version_doi":"10.13026/ckkp-6y19","is_latest_version":true,"short_description":"RadQA is an electronic health record question answering dataset containing clinical questions that can be answered using the Findings and Impressions sections of radiology reports","publish_date":"2022-12-09","title":"RadQA: A Question Answering Dataset to Improve Comprehension of Radiology Reports","abstract":"We present a radiology question answering dataset, RadQA, with 3074 questions posed against radiology reports and annotated with their corresponding answer spans (resulting in a total of 6148 question-answer evidence pairs) by physicians. The questions are manually created using the clinical referral section of the reports that take into account the actual information needs of ordering physicians and eliminate bias from seeing the answer context (and, further, organically create unanswerable questions). The answer spans are marked within the Findings and Impressions sections of a report. The dataset aims to satisfy the complex clinical requirements by including complete (yet concise) answer phrases (which are not just entities) that can span multiple lines. In published work, we conducted a thorough analysis of the proposed dataset by examining the broad categories of disagreement in annotation (providing insights on the errors made by humans) and the reasoning requirements to answer a question (uncovering the huge dependence on medical knowledge for answering the questions). In that work, the best-performing transformer language model achieved an F1 of 63.55 on the test set. However, the top human-level performance on this dataset is 90.31 (with an average human performance of 84.52), which demonstrates the challenging nature of RadQA that leaves ample scope for future method research.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":3914236,"compressed_storage_size":534058,"resource_type":"Database","access_policy":"Credentialed","topics":["question answering","clinical notes","radiology reports","machine reading comprehension","electronic health records"],"source_url":"https://physionet.org/content/radqa/1.0.0/"},{"public_project_uuid":"131879fd-5686-464d-be13-d335114f7a2d","slug":"kinecal","version":"1.0.2","core_doi":"10.13026/8bsb-ch59","version_doi":"10.13026/ky1q-3h26","is_latest_version":false,"short_description":"A dataset for balance falls-risk assessment and balance impairment analysis","publish_date":"2022-12-22","title":"KINECAL","abstract":"The field of human action recognition has made great strides in recent years, much helped by the availability of a wide variety of datasets that use Kinect to record human movement. Conversely, progress towards using Kinect in clinical practice has been hampered by the lack of appropriate data. In particular, datasets that contain clinically significant movements and appropriate metadata. This paper proposes a dataset to address this issue, namely KINECAL. It contains the recordings of 90 individuals carrying out 11 movements, commonly used in the clinical assessment of balance. The dataset contains relevant metadata, including clinical labelling, falls history and postural sway metrics. In addition, the recordings were made in both the lab and in informal settings.
","license":{"name":"Creative Commons Zero 1.0 Universal Public Domain Dedication"},"dua":null,"main_storage_size":192423502008,"compressed_storage_size":92063955575,"resource_type":"Database","access_policy":"Open","topics":["balance","postural sway","posturography","clinical tests","falls-risk","age-related changes"],"source_url":"https://physionet.org/content/kinecal/1.0.2/"},{"public_project_uuid":"97dbddcf-d6be-4f55-8a6d-05b18279ce45","slug":"mimic-iv-ecg","version":"0.1","core_doi":"10.13026/b95v-ff39","version_doi":"10.13026/tw25-ec93","is_latest_version":false,"short_description":"The MIMIC-IV ECG module contains approximately 800,000 diagnostic electrocardiograms across nearly 160,000 unique patients. These patients overlap with the patients from the MIMIC-IV Clinical Database.","publish_date":"2022-12-23","title":"MIMIC-IV-ECG - Diagnostic Electrocardiogram Matched Subset","abstract":"The MIMIC-IV-ECG module contains approximately 800,000 diagnostic electrocardiograms across nearly 160,000 unique patients. These diagnostic ECGs use 12 leads and are 10 seconds in length. They are sampled at 500 Hz. This subset contains all of the ECGs for patients who appear in the MIMIC-IV Clinical Database. When a cardiologist report is available for a given ECG, it is also provided. The patients in MIMIC-IV-ECG have been matched against the MIMIC-IV Clinical Database, making it possible to link to information across the MIMIC-IV modules.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":96912024711,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimic-iv-ecg/0.1/"},{"public_project_uuid":"4876b8e7-2ae7-4223-ac87-5a1a6e25cff8","slug":"mimic-iv-note","version":"2.1","core_doi":"10.13026/7qgp-kc16","version_doi":"10.13026/0p14-t007","is_latest_version":false,"short_description":"Deidentified free-text clinical notes for patients in the MIMIC-IV Clinical Database.","publish_date":"2023-01-05","title":"MIMIC-IV-Note: Deidentified free-text clinical notes","abstract":"The advent of large, open access text databases has driven advances in state-of-the-art model performance in natural language processing (NLP). The relatively limited amount of clinical data available for NLP has been cited as a significant barrier to the field's progress. Here we describe MIMIC-IV-Note: a collection of deidentified free-text clinical notes for patients included in the MIMIC-IV clinical database. MIMIC-IV-Note contains 357,289 deidentified discharge summaries from 161,403 patients admitted to the hospital and emergency department at the Beth Israel Deaconess Medical Center in Boston, MA, USA. The database also contains 2,471,881 deidentified radiology reports for 256,400 patients. All notes have had protected health information removed in accordance with the Health Insurance Portability and Accountability Act (HIPAA) Safe Harbor provision. All notes are linkable to MIMIC-IV providing important context to the clinical data therein. The database is intended to stimulate research in clinical natural language processing and associated areas.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1961480850,"compressed_storage_size":1961385798,"resource_type":"Database","access_policy":"Credentialed","topics":["clinical notes","mimic","critical care","electronic health record","natural language processing","deidentification"],"source_url":"https://physionet.org/content/mimic-iv-note/2.1/"},{"public_project_uuid":"37dc2b78-1ff6-45e2-8dd6-cb8da868113f","slug":"cardiac-accel-canine-porcine","version":"1.0.0","core_doi":"10.13026/hp2p-mz58","version_doi":"10.13026/5xmr-zd88","is_latest_version":true,"short_description":"The dataset contains data recorded from epicardially attached accelerometer to canines' and porcines' hearts. The data comprises of tri-axial acceleration signals, left ventricular pressure (LVP), rate of change of LVP and in some cases ECG as well.","publish_date":"2023-01-06","title":"Epicardially attached cardiac accelerometer data from canines and porcines","abstract":"The database contains acceleration signals from accelerometers attached directly on the epicardium in canines and porcines during various interventions that altered cardiac function and myocardial motion. The tri-axial accelerometer was attached in the anterior, apical region of the left ventricle (LV). The database contains a total of 289 recordings from 19 canines and 27 porcines with an average duration of 15 seconds, ranging from 5 to 30 seconds. Acceleration data was recorded simultaneously with ECG and LV pressure (LVP).
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":375420695,"compressed_storage_size":155206956,"resource_type":"Database","access_policy":"Open","topics":["heart","aortic valve","accelerometer","mitral valve"],"source_url":"https://physionet.org/content/cardiac-accel-canine-porcine/1.0.0/"},{"public_project_uuid":"7b357868-0981-4d4f-86d0-c6570296a753","slug":"mimic-iv-ed","version":"2.2","core_doi":"10.13026/1cjn-2370","version_doi":"10.13026/5ntk-km72","is_latest_version":true,"short_description":"A large database of emergency department admissions.","publish_date":"2023-01-06","title":"MIMIC-IV-ED","abstract":"MIMIC-IV-ED is a large, freely available database of emergency department (ED) admissions at the Beth Israel Deaconess Medical Center between 2011 and 2019. The database contains ~425,000 ED stays. Vital signs, triage information, medication reconciliation, medication administration, and discharge diagnoses are available. All data are deidentified to comply with the Health Information Portability and Accountability Act (HIPAA) Safe Harbor provision. MIMIC-IV-ED is intended to support a diverse range of education initiatives and research studies.","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":121978857,"compressed_storage_size":121933629,"resource_type":"Database","access_policy":"Credentialed","topics":["emergency department","ed","mimic","mimic-iv","electronic health record","emergency"],"source_url":"https://physionet.org/content/mimic-iv-ed/2.2/"},{"public_project_uuid":"cb81b1a5-9399-44a0-b004-d1c581afee2d","slug":"mimic-iv-note","version":"2.2","core_doi":"10.13026/7qgp-kc16","version_doi":"10.13026/1n74-ne17","is_latest_version":true,"short_description":"Deidentified free-text clinical notes for patients in the MIMIC-IV Clinical Database.","publish_date":"2023-01-06","title":"MIMIC-IV-Note: Deidentified free-text clinical notes","abstract":"The advent of large, open access text databases has driven advances in state-of-the-art model performance in natural language processing (NLP). The relatively limited amount of clinical data available for NLP has been cited as a significant barrier to the field's progress. Here we describe MIMIC-IV-Note: a collection of deidentified free-text clinical notes for patients included in the MIMIC-IV clinical database. MIMIC-IV-Note contains 331,794 deidentified discharge summaries from 145,915 patients admitted to the hospital and emergency department at the Beth Israel Deaconess Medical Center in Boston, MA, USA. The database also contains 2,321,355 deidentified radiology reports for 237,427 patients. All notes have had protected health information removed in accordance with the Health Insurance Portability and Accountability Act (HIPAA) Safe Harbor provision. All notes are linkable to MIMIC-IV providing important context to the clinical data therein. The database is intended to stimulate research in clinical natural language processing and associated areas.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1961480850,"compressed_storage_size":1961385799,"resource_type":"Database","access_policy":"Credentialed","topics":["critical care","mimic","natural language processing","clinical notes","deidentification","electronic health record"],"source_url":"https://physionet.org/content/mimic-iv-note/2.2/"},{"public_project_uuid":"53cd3164-37fb-40cb-8139-6aee03df73b0","slug":"mimiciv","version":"2.2","core_doi":"10.13026/07hj-2a80","version_doi":"10.13026/6mm1-ek67","is_latest_version":false,"short_description":"Large database of de-identified health information from patients admitted to Beth Israel Deaconess Medical Center","publish_date":"2023-01-06","title":"MIMIC-IV","abstract":"Retrospectively collected medical data has the opportunity to improve patient care through knowledge discovery and algorithm development. Broad reuse of medical data is desirable for the greatest public good, but data sharing must be done in a manner which protects patient privacy. The Medical Information Mart for Intensive Care (MIMIC)-III database provided critical care data for over 40,000 patients admitted to intensive care units at the Beth Israel Deaconess Medical Center (BIDMC). Importantly, MIMIC-III was deidentified, and patient identifiers were removed according to the Health Insurance Portability and Accountability Act (HIPAA) Safe Harbor provision. MIMIC-III has been integral in driving large amounts of research in clinical informatics, epidemiology, and machine learning. Here we present MIMIC-IV, an update to MIMIC-III, which incorporates contemporary data and improves on numerous aspects of MIMIC-III. MIMIC-IV adopts a modular approach to data organization, highlighting data provenance and facilitating both individual and combined use of disparate data sources. MIMIC-IV is intended to carry on the success of MIMIC-III and support a broad set of applications within healthcare.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":7688002262,"compressed_storage_size":7620008759,"resource_type":"Database","access_policy":"Credentialed","topics":["critical care","mimic","intensive care unit","machine learning"],"source_url":"https://physionet.org/content/mimiciv/2.2/"},{"public_project_uuid":"f0f9c4b8-09fd-4566-a5c7-930745c9837e","slug":"simultaneous-measurements","version":"1.0.2","core_doi":"10.13026/ddzr-dw82","version_doi":"10.13026/wce5-fj54","is_latest_version":true,"short_description":"Dataset to support comparison of usability and accuracy from simultaneous measurements collected from 13 subjects including five devices: NeXus-10 MKII, eMotion Faros 360°, Hexoskin Hx1, SOMNOTouch NIBP, Polar RS800 Multi.","publish_date":"2023-01-18","title":"Simultaneous physiological measurements with five devices at different cognitive and physical loads","abstract":"This open database contains physiological measurements from 13 adult and healthy subjects during a standardized experimental setup. The physiology was examined with five different devices (NeXus-10 MKII, eMotion Faros 360°, SOMNOtouch NIBP, Hexoskin Hx1, Polar RS800 Multi), which, in contrast to other studies, were recorded simultaneously. The experiment included a five minute baseline measurement, five minute walking on the treadmill, a five minute cognitive audio test, and a five minute uphill walking on the treadmill. Measurements included electrocardiography (ECG), photoplethysmography, accelerometry, oxygen saturation, respiration, heart rate, heart rate variability, and RR intervals sampled between 1 Hz and 8000 Hz. The dataset was originally generated during an investigation of functionality, accuracy, and usability of several ECG-measurement devices. This resource included the complete raw data files (EDF and HRM) and merged data with a manual expert annotation of heart beats and annotation files with information about the beginning of each experimental phase.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":1783802925,"compressed_storage_size":808909021,"resource_type":"Database","access_policy":"Open","topics":["movement","heart rate","ecg","photoplethysmogram","temperature","holter","accelerometer","noise","multiparameter","hrv","respiration"],"source_url":"https://physionet.org/content/simultaneous-measurements/1.0.2/"},{"public_project_uuid":"5612fcfa-5fb9-4ea5-b20a-9be9142bc4ab","slug":"zhejiang-ehr-critical-care","version":"1.0","core_doi":"10.13026/3h21-rc35","version_doi":"10.13026/901c-yv54","is_latest_version":true,"short_description":"Chinese critical care database from electronic healthcare records in a tertiary care medical center","publish_date":"2023-01-19","title":"Establishment of a Chinese critical care database from electronic healthcare records in a tertiary care medical center","abstract":"The medical specialty of critical care, or intensive care, provides emergency medical care to patients suffering from life-threatening complications and injuries. The medical specialty is featured by the generation of a huge amount of high-granularity data in routine practice. The data comprise hourly vital signs, ventilator waveforms, medical orders, and laboratory results. Currently, these data are well archived in the hospital information system for the primary purpose of routine clinical practice. However, data scientists have noticed that in-depth mining of such big data may provide insights into the pathophysiology of underlying diseases and healthcare practices. Clinical questions related to risk factors, predictive analytics, cost-effectiveness, and causal inference can be addressed with the critical care database. There have been several openly accessible critical care databases being established, which have generated hundreds of scientific outputs published scientific journals. However, such work is still in its infancy in China. China has a large patient population, which contributes to the generation of a large healthcare database in hospitals. The establishment and sharing of such a database can help to promote the open data science and discover novel scientific knowledge in a collaborative way. In this data descriptor article, we report the establishment of an openly accessible critical care database generated from hospital information system. The database comprises 8180 unique hospital admissions for 7638 individual patients from January 2012 to May 2022. The database contains 11 plain text relational tables that can be linked by hospital_ID.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":5549727049,"compressed_storage_size":511425158,"resource_type":"Database","access_policy":"Credentialed","topics":["critical care","database","china"],"source_url":"https://physionet.org/content/zhejiang-ehr-critical-care/1.0/"},{"public_project_uuid":"af628f26-ad6f-4d0f-9a34-b7a8147b3a25","slug":"bionlp-workshop-2023-task-1a","version":"1.0.0","core_doi":"10.13026/s8wk-ja78","version_doi":"10.13026/1z6g-ex18","is_latest_version":false,"short_description":"This is the data storage for BioNLP Workshop Shared Task 1A: Problem List Summarization.","publish_date":"2023-01-19","title":"BioNLP Workshop 2023 Shared Task 1A: Problem List Summarization","abstract":"Automatically summarizing patients’ main problems from the daily care notes in the electronic health record can help mitigate information and cognitive overload for clinicians and provide augmented intelligence via computerized diagnostic decision support at the bedside. The task of Problem List Summarization aims to generate a list of diagnoses and problems in a patient’s daily care plan using input from the provider’s progress notes during hospitalization.
\n\nThis task aims to promote NLP model development for downstream applications in diagnostic decision support systems that could improve efficiency and reduce diagnostic errors in hospital care. The task contains 768 hospital daily progress notes and 2783 diagnoses in the training set, and physicians will annotate a new set of 300 daily progress notes as the test set. The annotation methods and annotation quality have previously been reported.
\n\nThe dataset supports a more complex summarization task to generate a list of relevant diagnoses/problems given the information in the Subjective, Objective, and Assessment sections of the note. Only diagnoses/problems that are available in the progress note were labeled for the task.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":2124277,"compressed_storage_size":658926,"resource_type":"Challenge","access_policy":"Contributor Review","topics":["bionlp","electronic health record summarization","clinical natural language processing"],"source_url":"https://physionet.org/content/bionlp-workshop-2023-task-1a/1.0.0/"},{"public_project_uuid":"419a395b-04f7-44d1-8a54-cc413f295c2f","slug":"mimicel-ed","version":"1.1.0","core_doi":"10.13026/j8yc-gg94","version_doi":"10.13026/9x27-dd33","is_latest_version":false,"short_description":"MIMIC-IV Event Log for Emergency Department","publish_date":"2023-01-24","title":"MIMICEL: MIMIC-IV Event Log for Emergency Department","abstract":"In this work, we extract an event log from the MIMIC-IV-ED dataset by adopting a well-established event log generation methodology, and we name this event log MIMICEL. The data tables in the MIMIC-IV-ED dataset relate to each other based on the existing relational database schema, and each table records the individual activities of patients along their journey in the emergency department (ED). While the data tables in the MIMIC-IV-ED dataset catch snapshots of a patient journey in the ED, the extracted event log MIMICEL aims to capture an end-to-end patient journey process. This will enable us to analyse the existing patient flows, thereby improving the efficiency of an ED process.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":9213846704,"compressed_storage_size":503905250,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimicel-ed/1.1.0/"},{"public_project_uuid":"9c49b3d8-6850-44db-b167-9d0f5d45a182","slug":"wfdb-python","version":"4.1.0","core_doi":"10.13026/jt5e-gq15","version_doi":"10.13026/9njx-6322","is_latest_version":true,"short_description":"Tools for working with waveforms in Python.","publish_date":"2023-01-25","title":"Waveform Database Software Package (WFDB) for Python","abstract":"Physiological waveforms - such as electrocardiograms (ECG), electroencephalograms (EEG), electromyograms (EMG) - are generated during the course of routine care. These signals contain information that can be used to understand underlying conditions of health. Effective processing and analysis of physiological data requires specialized software. The waveform-database (WFDB) package for Python is a library of tools for reading, writing, and processing physiological signals and annotations, written in the Python programming language. Core components of the Python package are loosely based on the specifications of the original C-language WFDB software. We aim to implement as many of the core WFDB features as possible, with user-friendly APIs. Additional features are added over time.
","license":{"name":"MIT License"},"dua":null,"main_storage_size":246143429,"compressed_storage_size":104257524,"resource_type":"Software","access_policy":"Open","topics":["waveform","wfdb","python"],"source_url":"https://physionet.org/content/wfdb-python/4.1.0/"},{"public_project_uuid":"63493001-ff9b-44eb-a405-d4a4d6eab82f","slug":"clinical-t5","version":"1.0.0","core_doi":"10.13026/aw2e-he88","version_doi":"10.13026/rj8x-v335","is_latest_version":true,"short_description":"We train a T5-Base and T5-Large from scratch on MIMIC-III and MIMIC-IV. Additionally, we further pretrain T5-Base and SciFive on notes from MIMIC. We release these model weights on PhysioNet.","publish_date":"2023-01-25","title":"Clinical-T5: Large Language Models Built Using MIMIC Clinical Text","abstract":"Recent advances in scaling large language models (LLMs) has resulted in significant improvements over a number of natural language processing benchmarks. There has been some work to pretrain these language models over clinical text. These works demonstrate that training a language model using masked language modeling (MLM) on clinical notes is an effective technique for boosting performance on downstream tasks. All of these previous works use decoder-only architectures. We train 4 different clinical T5 models on the union of MIMIC-III and IV notes. Two of the models are initialized from previous T5-models (T5-base and SciFive). We additionally train a T5-Base and T5-Large model from scratch. These models should not be distributed to non-credentialed users. Research has shown that these language models have the potential to leak sensitive information. Due to this potential risk, we release the model weights under PhysioNet credentialed access.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":5636389253,"compressed_storage_size":5223086548,"resource_type":"Model","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/clinical-t5/1.0.0/"},{"public_project_uuid":"d9838c5a-91f4-48d0-a52e-d6c3de3d3b86","slug":"pressure-flow-circum-cpap","version":"1.0.0","core_doi":"10.13026/wmzm-d487","version_doi":"10.13026/25q9-y416","is_latest_version":true,"short_description":"Dataset of pressure, flow, and dynamic abdominal and chest circumference for healthy people breathing with CPAP. Data was collected with PEEP settings of 0 (ZEEP), 4, and 8cmH2O at normal/resting, panting/short and deep/long breath patterns/rates.","publish_date":"2023-01-25","title":"Pressure, flow, and dynamic thoraco-abdominal circumferences data for adults breathing under CPAP therapy","abstract":"A respiratory pressure, flow, and dynamic circumference (chest and abdominal) dataset was collected from 30 healthy adults at the University of Canterbury under Human Research Ethics Committee (HREC) ethics. Pressure and flow data was collected using a bi-directional Venturi, with data combined in post-collection processing. Dynamic circumferences were measured using inelastic band tape measures attached to rotary encoders capturing expansion and contraction. Data was collected at zero end-expiratory pressure (ZEEP) and with continuous positive airway pressure (CPAP) ventilation, at positive end-expiratory pressure (PEEP) settings of 4 and 8 cmH2O. Data was also collected over three breath types, in which subjects were instructed to breathe normally, to pant as fast as possible, and to breathe as deeply as possible. Recording times were set to 65 seconds for normal and deep breathing, and 35 seconds for panting trials. Subjects self-determined rest time between trials. A questionnaire was used to collect relevant demographic data (sex, height, age, weight, and history of smoking or asthma). The dataset was collected to inform and validate model-based respiratory monitoring systems, with the ultimate goal of shifting respiratory testing from specialist clinical settings to primary care and home settings
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":319250373,"compressed_storage_size":83748781,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/pressure-flow-circum-cpap/1.0.0/"},{"public_project_uuid":"7538a155-7ca4-4c4d-95e7-5650f09759f6","slug":"mimic-iv-demo","version":"2.2","core_doi":"10.13026/ng9m-3n32","version_doi":"10.13026/dp1f-ex47","is_latest_version":true,"short_description":"An openly available subset of patients in the MIMIC-IV database.","publish_date":"2023-01-31","title":"MIMIC-IV Clinical Database Demo","abstract":"The Medical Information Mart for Intensive Care (MIMIC)-IV database is comprised of deidentified electronic health records for patients admitted to the Beth Israel Deaconess Medical Center. Access to MIMIC-IV is limited to credentialed users. Here, we have provided an openly-available demo of MIMIC-IV containing a subset of 100 patients. The dataset includes similar content to MIMIC-IV, but excludes free-text clinical notes. The demo may be useful for running workshops and for assessing whether the MIMIC-IV is appropriate for a study before making an access request.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":16224447,"compressed_storage_size":16189661,"resource_type":"Database","access_policy":"Open","topics":["mimic","critical care","electronic health record"],"source_url":"https://physionet.org/content/mimic-iv-demo/2.2/"},{"public_project_uuid":"564d1298-36eb-415e-97d6-d78b6cf6dce1","slug":"mimic-iv-ecg","version":"0.2","core_doi":"10.13026/b95v-ff39","version_doi":"10.13026/a1j7-ky04","is_latest_version":false,"short_description":"The MIMIC-IV ECG module contains approximately 800,000 diagnostic electrocardiograms across nearly 160,000 unique patients. These patients overlap with the patients from the MIMIC-IV Clinical Database.","publish_date":"2023-02-08","title":"MIMIC-IV-ECG - Diagnostic Electrocardiogram Matched Subset","abstract":"The MIMIC-IV-ECG module contains approximately 800,000 diagnostic electrocardiograms across nearly 160,000 unique patients. These diagnostic ECGs use 12 leads and are 10 seconds in length. They are sampled at 500 Hz. This subset contains all of the ECGs for patients who appear in the MIMIC-IV Clinical Database. When a cardiologist report is available for a given ECG, it is also provided. The patients in MIMIC-IV-ECG have been matched against the MIMIC-IV Clinical Database, making it possible to link to information across the MIMIC-IV modules.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":96912024711,"compressed_storage_size":36271897681,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimic-iv-ecg/0.2/"},{"public_project_uuid":"c63699e2-8200-4041-a0ba-8859b9bdc634","slug":"mimic-iv-ed-demo","version":"2.2","core_doi":"10.13026/edf9-9g37","version_doi":"10.13026/jzz5-vs76","is_latest_version":true,"short_description":"An openly available subset of the MIMIC-IV-ED database","publish_date":"2023-02-08","title":"MIMIC-IV-ED Demo","abstract":"MIMIC-IV-ED is a publicly accessible database of over 400,000 emergency department (ED) admissions to the Beth Israel Deaconess Medical Center between 2011 and 2019. Access to MIMIC-IV-ED requires registration on PhysioNet, identity verification, completion of human participant training, and signing of a data use agreement. Here, we have provided an openly available demo of MIMIC-IV-ED containing a subset of 100 patients. Being a derivative dataset, the MIMIC-IV-ED demo is deidentified according to the same standard as MIMIC-IV-ED. The demo is intended to support workshops, educational material, exploration of MIMIC-IV-ED, and other aims enabled by an openly available dataset.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":114530,"compressed_storage_size":97782,"resource_type":"Database","access_policy":"Open","topics":["mimic","emergency department"],"source_url":"https://physionet.org/content/mimic-iv-ed-demo/2.2/"},{"public_project_uuid":"50e812ed-6dd2-4c64-84c8-1e0c9cd61cb8","slug":"chest-x-ray-segmentation","version":"1.0.0","core_doi":"10.13026/pa5n-mc06","version_doi":"10.13026/9cy4-f535","is_latest_version":true,"short_description":"CXLSeg dataset: Chest X-ray with Lung Segmentation, a comparatively large dataset of segmented Chest X-ray radiographs based on the MIMIC-CXR dataset. This contains segmentation results of 243,324 frontal view images and corresponding masks.","publish_date":"2023-02-08","title":"Chest X-ray Dataset with Lung Segmentation","abstract":"Chest X-ray(CXR) images are prominent among medical images and are commonly administered in emergency diagnosis and treatment corresponding to cardiac and respiratory diseases. Though there are robust solutions available for medical diagnosis, validation of artificial intelligence (AI) in radiology is still questionable. Segmentation is pivotal in chest radiographs that aid in improvising the existing AI-based medical diagnosis process. We provide the CXLSeg dataset: Chest X-ray with Lung Segmentation, a comparatively large dataset of segmented Chest X-ray radiographs based on the MIMIC-CXR dataset, a popular CXR image dataset. The dataset contains segmentation results of 243,324 frontal view images of the MIMIC-CXR dataset and corresponding masks. Additionally, this dataset can be utilized for computer vision-related deep learning tasks such as medical image classification, semantic segmentation and medical report generation. Models using segmented images yield better results since only the features related to the important areas of the image are focused. Thus images of this dataset can be manipulated to any visual feature extraction process associated with the original MIMIC-CXR dataset and enhance the results of the published or novel investigations. Furthermore, masks provided by this dataset can be used to train segmentation models when combined with the MIMIC-CXR-JPG dataset. The SA-UNet model achieved a 96.80% in dice similarity coefficient and 91.97% in IoU for lung segmentation using CXLSeg.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":4626095644,"compressed_storage_size":4362042641,"resource_type":"Database","access_policy":"Credentialed","topics":["medical reports","mimic-cxr","u-net","chest x-ray","chest radiographs","segmentation"],"source_url":"https://physionet.org/content/chest-x-ray-segmentation/1.0.0/"},{"public_project_uuid":"18602ed5-f7a2-4dab-919d-f91ead920b03","slug":"annotation-opioid-use-notes","version":"1.0.0","core_doi":"10.13026/kdap-sg35","version_doi":"10.13026/tbb4-t319","is_latest_version":true,"short_description":"The database contains a corpus of annotated data from the MIMIC-III Critical Care Database from a study that aimed to develop and apply an annotation schema to characterize opioid use disorder and related contextual factors.","publish_date":"2023-02-08","title":"Annotation dataset of problematic opioid use and related contexts from MIMIC-III Critical Care Database discharge summaries","abstract":"Opioid use disorder (OUD) is underdiagnosed in health system settings, limiting research on OUD using electronic health records (EHRs). Medical encounter notes can enrich structured EHR data with documented signs and symptoms of OUD and social risks and behaviors. To capture this information at scale, natural language processing tools must be developed and evaluated. We conducted a pilot study that aimed to 1) develop and apply an annotation schema to deeply characterize OUD and related clinical, behavioral, and environmental factors; and 2) automate the annotation schema using machine learning and deep learning-based approaches. De-identified patient data for this study included hospital discharge summaries of patients with International Classification of Diseases (ICD-9) OUD diagnostic codes, obtained from the MIMIC-III Critical Care Database. We developed an annotation schema to characterize problematic opioid use, identify individuals with potential OUD, and provide psychosocial context. The final annotation schema contained 33 classes. Two annotators reviewed discharge summaries from a random sample of 100 of these patients. The first corpus of 40 patients was reviewed by both annotators. We achieved moderate inter-annotator agreement, with F1-scores across all classes increasing from 48% to 66%. The second corpus of 60 patients was reviewed by a single annotator. The shared database contains the resulting 3,270 annotations with the note identifier, span offset with accompanying text snippet, and class assignments and may be useful to future development of natural language processing systems related to OUD.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":605002,"compressed_storage_size":92922,"resource_type":"Database","access_policy":"Credentialed","topics":["clinical notes","substance use","natural language processing","opioid use disorder"],"source_url":"https://physionet.org/content/annotation-opioid-use-notes/1.0.0/"},{"public_project_uuid":"c20dc22c-e19f-48bf-8b48-c7debc32528b","slug":"in-gauge-and-en-gage","version":"1.0.0","core_doi":"10.13026/zcpz-q137","version_doi":"10.13026/srm3-7z33","is_latest_version":true,"short_description":"The project aims to understand occupants’ behaviour, engagement, emotion, and comfort indoors with heterogeneous sensors and wearables.","publish_date":"2023-02-13","title":"In-Gauge and En-Gage: Understanding Occupants' Behaviour, Engagement, Emotion, and Comfort Indoors with Heterogeneous Sensors and Wearables","abstract":"We conducted a field study at a K-12 private school in the suburbs of Melbourne, Australia. The datasets contained two elements: (1) In-Gauge dataset: we conducted a 5-month longitudinal field study using two outdoor weather stations, as well as indoor weather stations in 17 classrooms and temperature sensors on the vents of occupant-controlled room air-conditioners; these were collated into individual datasets for each classroom at a 5-minute logging frequency, including additional data on occupant presence. (2) En-Gage dataset: we tracked 23 students and 6 teachers in a 4-week cross-sectional study En-Gage, using wearable sensors to log physiological data (electrodermal activity, heart rate, blood column pulse, skin temperature, 3-axis acceleration), as well as daily surveys to query the occupants' thermal comfort, learning engagement, emotions and seating behaviours. Overall, the combined dataset could be used to analyse the relationships between indoor/outdoor climates and students' behaviours/mental states on campus, which provide opportunities for the future design of intelligent feedback systems to benefit both students and staff.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":13757978126,"compressed_storage_size":5448876957,"resource_type":"Database","access_policy":"Open","topics":["heart rate","environmental sensing","thermal comfort modelling","physiological signals","electrodermal activity","occupant behaviour sensing","emotion sensing","human behavoural modelling","smart building"],"source_url":"https://physionet.org/content/in-gauge-and-en-gage/1.0.0/"},{"public_project_uuid":"07965158-121e-4bb2-8e8a-10503fdea6ed","slug":"sicdb","version":"1.0.5","core_doi":"10.13026/ezs8-6v88","version_doi":"10.13026/67ww-6m34","is_latest_version":false,"short_description":"The SICdb dataset, version 1.0.5 contains 27386 admissions to an ICU in an Austrian tertiary care institution.","publish_date":"2023-02-14","title":"Salzburg Intensive Care database (SICdb), a freely accessible intensive care database","abstract":"The SICdb dataset offers insights into over 27 thousand intensive care admissions, including therapies and data on preceding surgeries. Data were collected between 2013 and 2021 from four different intensive care units at the University Hospital Salzburg, having more than 3 thousand intensive care admissions per year on 43 beds. The dataset is deidentified and contains, amongst others, case information, vital signs, laboratory results and medication data. SICdb provides both aggregated once-per-hour and highly granular once-per-minute data, making it suitable for computational and machine learning-based research.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":2086212223,"compressed_storage_size":2082990360,"resource_type":"Database","access_policy":"Contributor Review","topics":["machine learning","clinical","critical care","intensive care"],"source_url":"https://physionet.org/content/sicdb/1.0.5/"},{"public_project_uuid":"75150baf-2e67-48ad-8eaf-33f6173659e7","slug":"i-care","version":"1.0","core_doi":"10.13026/sd64-2856","version_doi":"10.13026/rjbz-cq89","is_latest_version":false,"short_description":"The clinical and EEG data for this dataset originates from seven academic hospitals in the U.S. and Europe led by investigators part of the International Cardiac Arrest REsearch consortium (I-CARE).","publish_date":"2023-02-22","title":"I-CARE: International Cardiac Arrest REsearch consortium Database","abstract":"The International Cardiac Arrest REsearch consortium (I-CARE) Database includes baseline clinical information and continuous electroencephalography (EEG) recordings from comatose patients with cardiac arrest who were admitted to an intensive care unit in seven academic hospitals in the U.S. and Europe. Patients were monitored with 18 bipolar EEG channels from hours to several days for the diagnosis of seizures as well as for neurological prognostication. Long-term neurological function was determined using the Cerebral Performance Category scale.
","license":{"name":"Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":25440223598,"compressed_storage_size":18632148751,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/i-care/1.0/"},{"public_project_uuid":"5354479b-807b-4c92-83fb-32fd4926aaee","slug":"image-embeddings-mimic-cxr","version":"1.0","core_doi":"10.13026/t3kp-ve72","version_doi":"10.13026/pxc2-vx69","is_latest_version":true,"short_description":"This database contains compact information-rich embeddings of the MIMIC-CXR Database v2.0.0 using the CXR Foundation API v1.0.","publish_date":"2023-02-22","title":"Generalized Image Embeddings for the MIMIC Chest X-Ray dataset","abstract":"This database contains image embeddings from the DICOM images of the MIMIC V2.0.0 Chest X-Ray database, generated using the approach described in “Simplified Transfer Learning for Chest Radiography Models Using Less Data”. The image embeddings are compact, information-rich numerical vectors that can be used for classification tasks. Given the nature of the embeddings, less compute resources and even less data are necessary to achieve the same classification performance as the full CXR image in training a model.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1422718023,"compressed_storage_size":1408110263,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/image-embeddings-mimic-cxr/1.0/"},{"public_project_uuid":"c925e3a9-133c-416a-844a-f1b6f4b145b4","slug":"mimic-ii","version":"2.6.0","core_doi":"10.13026/qeq9-tw98","version_doi":"10.13026/fxn0-mk84","is_latest_version":true,"short_description":"Electronic health record data collected from >30,000 patients admitted to ICUs at a single tertiary care hospital.","publish_date":"2011-04-25","title":"MIMIC-II Clinical Database","abstract":"MIMIC-II documents a diverse and large population of intensive care unit patient stays and contains comprehensive and detailed clinical data, including physiological waveforms and minute-by-minute trends for a subset of records. It establishes a unique public-access resource for critical care research, supporting a diverse range of analytic studies spanning epidemiology, clinical decision-rule development, and electronic tool development. The MIMIC-II Clinical Database, although de-identified, still contains detailed information regarding the clinical care of patients, and must be treated with appropriate care and respect.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":2595,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":["ehr","mimic-ii","icu","bidmc"],"source_url":"https://physionet.org/content/mimic-ii/2.6.0/"},{"public_project_uuid":"324db101-1263-4eb6-a5cf-ab7eaff78753","slug":"mimicel-ed","version":"2.0.0","core_doi":"10.13026/j8yc-gg94","version_doi":"10.13026/zd69-tq37","is_latest_version":false,"short_description":"MIMIC-IV Event Log for Emergency Department","publish_date":"2023-03-06","title":"MIMICEL: MIMIC-IV Event Log for Emergency Department","abstract":"In this work, we extract an event log from the MIMIC-IV-ED dataset by adopting a well-established event log generation methodology, and we name this event log MIMICEL. The data tables in the MIMIC-IV-ED dataset relate to each other based on the existing relational database schema, and each table records the individual activities of patients along their journey in the emergency department (ED). While the data tables in the MIMIC-IV-ED dataset catch snapshots of a patient journey in the ED, the extracted event log MIMICEL aims to capture an end-to-end process of the patient journey. This will enable us to analyse the existing patient flows, thereby improving the efficiency of an ED process.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":9559619081,"compressed_storage_size":508003982,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimicel-ed/2.0.0/"},{"public_project_uuid":"c168e641-0193-4516-9b29-694b7735fcf9","slug":"big-ideas-glycemic-wearable","version":"1.1.0","core_doi":"10.13026/w591-tp72","version_doi":"10.13026/e827-h193","is_latest_version":false,"short_description":"Glucose measurements and wrist-worn wearable sensor data from highnormoglycemic participants.","publish_date":"2023-03-06","title":"BIG IDEAs Lab Glycemic Variability and Wearable Device Data","abstract":"This study aimed to determine the feasibility and effectiveness of wearable devices in detecting early physiological changes prior to the development of prediabetes [1-3]. The study generated digital biomarkers for remote, mHealth-based prediabetes and hyperglycemia risk to classify which individuals should undergo further clinical testing. The primary inclusion criteria were subjects aged 35-65 years, inclusive, including only post-menopausal females, with a point of care A1C measurement between 5.2-6.4%, inclusive. Blood was collected during the study for measurement of glucose, hemoglobin A1C, lipoproteins, and triglycerides. Participants wore a Dexcom 6 continuous glucose monitor (CGM) and an Empatica E4 wristband for 10 days while receiving a standardized breakfast meal every other day. At the end of the 10 days, the participant returned to the clinic for an oral glucose tolerance test (OGTT). Research data collected includes physiological measurements from wearable devices such as heart rate, accelerometry, and electrodermal conductance.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":36649629220,"compressed_storage_size":5016598065,"resource_type":"Database","access_policy":"Open","topics":["biomedical engineering","biomarkers","pre-diabetes"],"source_url":"https://physionet.org/content/big-ideas-glycemic-wearable/1.1.0/"},{"public_project_uuid":"c0f1fe63-5856-4771-bede-70e3d48a3786","slug":"brazilian-ophthalmological","version":"1.0.0","core_doi":"10.13026/z9zv-g542","version_doi":"10.13026/xcxw-8198","is_latest_version":false,"short_description":"This is the first Brazilian Multilabel Ophthalmological Dataset with demographic information and retinal photos labeled images according to anatomical parameters, quality control, and presumed diagnosis.","publish_date":"2023-03-08","title":"A Brazilian Multilabel Ophthalmological Dataset (BRSET)","abstract":"The Brazilian Multilabel Ophthalmological Dataset (BRSET) is a multi-labeled ophthalmological dataset designed to improve scientific community development and validate machine learning models. In ophthalmology, ancillary exams support medical decisions and can be used to develop algorithms; however, the availability and representativeness of ophthalmological datasets are limited. This dataset consists of 16,266 images from 8,524 Brazilian patients. Demographics, macula, optic disc, and vessels anatomical parameters, focus, illumination, image field, and artifacts as quality control, and multi-labels are included alongside color fundus retinal photos. This dataset enables computer vision models to predict demographic characteristics and multi-label disease classification using retinal fundus photos.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":16711707399,"compressed_storage_size":13072442102,"resource_type":"Database","access_policy":"Credentialed","topics":["ophthalmology","retina","dataset"],"source_url":"https://physionet.org/content/brazilian-ophthalmological/1.0.0/"},{"public_project_uuid":"7d5303a0-1bd2-40ee-a657-5bd47b5acb56","slug":"script-carpediem-dataset","version":"1.1.0","core_doi":"10.13026/1qke-qx33","version_doi":"10.13026/5phr-4r89","is_latest_version":false,"short_description":"SCRIPT seeks to delineate the host/pathogen interactions during pneumonia using multiomic analysis of bronchoalveolar lavage fluid joined with clinical data and physician adjudication.","publish_date":"2023-03-13","title":"SCRIPT CarpeDiem Dataset: demographics, outcomes, and per-day clinical parameters for critically ill patients with suspected pneumonia","abstract":"Traditional approaches to analyzing episodes of patient care in the ICU examine features on presentation and outcomes on discharge, collapsing the numerous events that happen during a patient’s stay and ignoring intercurrent ICU complications or interventions. With CarpeDiem, we aimed to examine clinical features on a per-day basis as done during the common practice of daily physician rounds. The Successful Clinical Response in Pneumonia Therapy (SCRIPT) CarpeDiem Dataset features 12,495 patient-ICU-days from 585 patients enrolled in the SCRIPT study between June 2018 and March 2022, all of whom were critically ill patients on mechanical ventilation suspected of having pneumonia who underwent a bronchoalveolar lavage as part of routine clinical care. Each patient has demographics, pneumonia episode information as adjudicated by a panel of critical care physicians, outcomes, and 44 clinical parameters for which data are present including vital signs, laboratory parameters, and mechanical support devices. Data have been deidentified per Health Insurance Portability and Accountability Act (HIPAA) Safe Harbor rules. This dataset combines expert clinician adjudication with per-day granular information to provide a unique tool to examine the clinical courses of critically ill patients with pneumonia. We use this dataset in a manuscript examining the contribution of ventilator-associated pneumonia to patient outcomes (see Usage section for details) and share it for others to work with.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":11574031,"compressed_storage_size":2338190,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/script-carpediem-dataset/1.1.0/"},{"public_project_uuid":"fa87da82-fa95-4960-81ef-59dc86bcf88e","slug":"globem","version":"1.1","core_doi":"10.13026/889w-p048","version_doi":"10.13026/r9s1-s711","is_latest_version":true,"short_description":"GLOBEM datasets contain the first released multi-year mobile and wearable sensing datasets from 2018 to 2021, containing 705 person-years and 497 unique participants.","publish_date":"2023-03-14","title":"GLOBEM Dataset: Multi-Year Datasets for Longitudinal Human Behavior Modeling Generalization","abstract":"We present the first multi-year mobile sensing datasets. Our multi-year data collection studies span four years (10 weeks each year, from 2018 to 2021). The four datasets contain data collected from 705 person-years (497 unique participants) with diverse racial, ability, and immigrant backgrounds. Each year, participants would install a mobile app on their phones and wear a fitness tracker. The app and wearable device passively track multiple sensor streams in the background 24×7, including location, phone usage, calls, Bluetooth, physical activity, and sleep behavior. In addition, participants completed weekly short surveys and two comprehensive surveys on health behaviors and symptoms, social well-being, emotional states, mental health, and other metrics. Our dataset analysis indicates that our datasets capture a wide range of daily human routines, and reveal insights between daily behaviors and important well-being metrics (e.g., depression status). We envision our multi-year datasets can support the ML community in developing generalizable longitudinal behavior modeling algorithms.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":3127170144,"compressed_storage_size":877632295,"resource_type":"Database","access_policy":"Credentialed","topics":["human behavior modeling","ubiquitous computing","well-being","health","passive mobile sensing"],"source_url":"https://physionet.org/content/globem/1.1/"},{"public_project_uuid":"230901c7-26f6-40e6-ba8d-a04e8e318874","slug":"ms-cxr-t","version":"1.0.0","core_doi":"10.13026/ga7g-1614","version_doi":"10.13026/pg10-j984","is_latest_version":true,"short_description":"The MS-CXR-T is a multimodal benchmark that enhances the MIMIC-CXR v2 dataset by including expert-verified annotations. Its goal is to evaluate biomedical visual-language processing models in terms of temporal semantics extracted from image and text.","publish_date":"2023-03-18","title":"MS-CXR-T: Learning to Exploit Temporal Structure for Biomedical Vision-Language Processing","abstract":"MS-CXR-T is a multi-modal benchmark dataset for evaluating biomedical vision-language processing (VLP) models on two distinct temporal tasks in radiology: image classification and sentence similarity. The former comprises multi-image frontal chest X-rays with ground-truth labels (N=1326) across 5 findings, with classes corresponding to 3 states of disease progression for each finding: {'Improving', 'Stable', 'Worsening'}, expanding on the Chest ImaGenome progression dataset. The latter quantifies the temporal-semantic similarity of text embeddings extracted from pairs of sentences (N=361). The pairs can be either paraphrases or contradictions in terms of disease progression. The data for both tasks was manually annotated and reviewed by a board-certified radiologist. The dataset provides researchers an opportunity to evaluate both image and text models on these biomedical temporal tasks and reproduce experiments reported in the corresponding literature.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":263165,"compressed_storage_size":95954,"resource_type":"Database","access_policy":"Credentialed","topics":["multimodal","cxr","vision-language processing","radiology","chest x-ray","disease progression"],"source_url":"https://physionet.org/content/ms-cxr-t/1.0.0/"},{"public_project_uuid":"96e1db00-27e5-4cbf-a657-1f80ef64d89f","slug":"mimic-eye-multimodal-datasets","version":"1.0.0","core_doi":"10.13026/kr1c-v245","version_doi":"10.13026/pc72-as03","is_latest_version":true,"short_description":"MIMIC-Eye: Integrating MIMIC Datasets with REFLACX and Eye Gaze for Multimodal Deep Learning Applications","publish_date":"2023-03-23","title":"MIMIC-Eye: Integrating MIMIC Datasets with REFLACX and Eye Gaze for Multimodal Deep Learning Applications","abstract":"Deep learning technologies have been widely adopted in medical imaging due to their ability to extract features from images and make accurate diagnoses automatically. Medical imaging technologies are particularly useful because they can be trained to detect subtle differences in images that are hard to detect for human radiologists. In the real world, radiologists must rely on various types of patient information to assess medical images confidently. However, most DL applications in medical imaging only utilize image data, mainly because the literature on medical datasets combining different data modalities is scarce. In this study, we present MIMIC-EYE, a dataset that encompasses a comprehensive integration of several datasets related to MIMIC. This dataset includes a comprehensive range of patient information, including medical images and reports (MIMIC CXR and MIMIC JPG), clinical data (MIMIC IV ED), a detailed account of the patient's hospital journey (MIMIC IV), and eye tracking data containing gaze information and pupil dilations together with image annotations (REFLACX and EYE GAZE). Integrating eye tracking data with the various MIMIC modalities may provide a more comprehensive understanding of radiologists' visual search behavior patterns and facilitate the development of more robust, accurate, and reproducible deep-learning models for medical imaging diagnosis.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":38920098062,"compressed_storage_size":25104557191,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/mimic-eye-multimodal-datasets/1.0.0/"},{"public_project_uuid":"54c02d24-6446-4544-afec-d716ccc88895","slug":"consumer-grade-wearables","version":"1.0.0","core_doi":"10.13026/21mx-7p04","version_doi":"10.13026/5f6t-b637","is_latest_version":true,"short_description":"Physiological data captured in experimental condition by three wearable devices.","publish_date":"2023-03-31","title":"CogWear: Can we detect cognitive effort with consumer-grade wearables?","abstract":"The CogWear database consists of two datasets: pilot and survey gamification experiment. The pilot dataset includes physiological signals from 11 volunteers captured by three wearable devices (Empatica E4 watch, Samsung Galaxy Watch4 and Muse S EEG headband) during cognitively demanding task (Stroop test) and baseline condition (no engagement in any activity). The pilot dataset was collected to evaluate the experimental setup and the signal captured by the wearables. In particular, can we detect when people perform a cognitively demanding task using data from consumer-grade smartwatches. The survey gamification experiment dataset includes physiological signal collected from 13 volunteers by three wearable devices (Empatica E4 watch, Samsung Galaxy Watch4 and Muse S EEG headband) during two sessions. In both sessions participants, as in the pilot study, engaged in a cognitively demanding task then rested and did not perform any tasks and finally filled in four mobile surveys related to work and wellbeing (GRIT, PANAS, HAPPINES and NFR). The mobile surveys were either gamified or not. The volunteers were divided into two groups: group 1 filled in gamified surveys in the first session and non-gamified surveys in the second session, group 2 had the opposite order. There was a two-week break time between the sessions. This dataset was captured to investigate whether simple gamification has an impact on the cognitive burden of filling in the surveys.
\n","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":2937057821,"compressed_storage_size":187856113,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/consumer-grade-wearables/1.0.0/"},{"public_project_uuid":"9b686630-c50f-4bbb-99b4-fe0e512e6743","slug":"scg-rhc-wearable-database","version":"1.0.0","core_doi":"10.13026/0mb3-4c08","version_doi":"10.13026/133d-pk11","is_latest_version":true,"short_description":"This is the first public dataset that contains simultaneous recordings of Right Heart Catheter data (pressure) and chest-worn wearable patch data (electrocardiogram and seismocardiogram signals).","publish_date":"2023-03-31","title":"SCG-RHC: Wearable Seismocardiogram Signal and Right Heart Catheter Database","abstract":"We are providing an open access dataset of 73 patients to study better ways to monitor changes in pulmonary arterial pressure (PAP) and pulmonary capillary wedge pressure (PCWP) using a small, non-invasive, chest-worn wearable patch. The ground truth PAP and PCWP were measured using pulmonary artery wedge catheter during the standard Right Heart Catheter (RHC) procedure, and they were modulated by infusing systemic vasodilators. The wearable data including electrocardiogram (ECG) and tri-axial seismocardiogram (SCG) were also measured simultaneously and time synchronized to the ground truth data. Besides waveform records, this dataset also contains demographic data such as sex, age, weight, height, heart failure status, etc.) and other RHC values such as right atrial pressure, right ventricular pressure, stroke volume, and cardiac output.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":12501195760,"compressed_storage_size":8510855981,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/scg-rhc-wearable-database/1.0.0/"},{"public_project_uuid":"f6e617fd-2358-406a-8dfb-328b62f9aeef","slug":"bionlp-workshop-2023-task-1a","version":"1.1.0","core_doi":"10.13026/s8wk-ja78","version_doi":"10.13026/1crd-qk14","is_latest_version":false,"short_description":"This is the data storage for BioNLP Workshop Shared Task 1A: Problem List Summarization.","publish_date":"2023-04-13","title":"BioNLP Workshop 2023 Shared Task 1A: Problem List Summarization","abstract":"Automatically summarizing patients’ main problems from the daily care notes in the electronic health record can help mitigate information and cognitive overload for clinicians and provide augmented intelligence via computerized diagnostic decision support at the bedside. The task of Problem List Summarization aims to generate a list of diagnoses and problems in a patient’s daily care plan using input from the provider’s progress notes during hospitalization.
\n\nThis task aims to promote NLP model development for downstream applications in diagnostic decision support systems that could improve efficiency and reduce diagnostic errors in hospital care. The task contains 768 hospital daily progress notes and 2783 diagnoses in the training set, and a new set of 237 daily progress notes are recently annotated as the test set. The annotation methods and annotation quality have previously been reported.
\n\nThe dataset supports a more complex summarization task to generate a list of relevant diagnoses/problems given the information in the Subjective, Objective, and Assessment sections of the note. Only diagnoses/problems that are available in the progress note were labeled for the task.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":2843706,"compressed_storage_size":872214,"resource_type":"Challenge","access_policy":"Contributor Review","topics":["electronic health record summarization","bionlp","clinical natural language processing"],"source_url":"https://physionet.org/content/bionlp-workshop-2023-task-1a/1.1.0/"},{"public_project_uuid":"d41caa4b-42dc-47c6-b4aa-aa0618ca2561","slug":"ptb-xl-plus","version":"1.0.0","core_doi":"10.13026/nqsf-pc74","version_doi":"10.13026/m5qc-8m53","is_latest_version":false,"short_description":"ECG feature dataset accompanying the PTB-XL ECG dataset","publish_date":"2023-04-25","title":"PTB-XL+, a comprehensive electrocardiographic feature dataset","abstract":"The PTB-XL+ dataset is a comprehensive feature dataset that supplements the PTB-XL ECG dataset. It includes ECG features extracted via two commercial and one open-source algorithm in a harmonized format as well as median beats or, where applicable, fiducial points extracted by the respective algorithms. In addition, it provides automatic diagnosis statements from one commercial ECG analysis algorithm at different processing levels that are ready to be used for training and evaluation of machine learning models.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":2182061968,"compressed_storage_size":733758236,"resource_type":"Database","access_policy":"Open","topics":["electrocardiography","ecg","ptb-xl","ptb"],"source_url":"https://physionet.org/content/ptb-xl-plus/1.0.0/"},{"public_project_uuid":"b4dd350b-2d0d-4bca-8f70-3f57c1e2b145","slug":"physiotag","version":"1.0.0","core_doi":"10.13026/x3jv-ch75","version_doi":"10.13026/g06j-3612","is_latest_version":true,"short_description":"Platform for collaborative and interactive annotation of physiological waveform data.","publish_date":"2023-04-25","title":"PhysioTag: An Open-Source Platform for Collaborative Annotation of Physiological Waveforms","abstract":"To develop robust algorithms for automated diagnosis of medical conditions such as cardiac arrhythmias, researchers require large collections of data with human expert annotations. Currently, there is a lack of accessible, open-source platforms for human experts to collaboratively develop these annotated datasets through a web interface. In this work, we developed a flexible, generalizable, web-based framework to enable multiple users to create and share annotations on multi-channel physiological waveforms. The software is simple to install and offers a range of features, including: user management and task customization; a programmatic interface for data import and export; and a leaderboard for annotation progress tracking.
","license":{"name":"MIT License"},"dua":null,"main_storage_size":59526374,"compressed_storage_size":47668375,"resource_type":"Software","access_policy":"Open","topics":["annotation"],"source_url":"https://physionet.org/content/physiotag/1.0.0/"},{"public_project_uuid":"2c470b5c-4491-4eca-aefd-531f83a298d4","slug":"medical-ai-research-foundation","version":"1.0.0","core_doi":"10.13026/psq3-vj24","version_doi":"10.13026/grp0-z205","is_latest_version":true,"short_description":"Medical AI Research Foundations is a repository of medical foundation models.","publish_date":"2023-04-26","title":"Medical AI Research Foundations: A repository of medical foundation models","abstract":"Medical AI Research Foundations is a repository of open-source medical foundation models. With this collection of non-diagnostic models, APIs, and resources like code and data, researchers and developers can accelerate their medical AI research. This is a clear unmet need as currently there is no central resource today that developers and researchers can leverage to build medical AI and as such, this has slowed down both research and translation efforts. Our goal is to democratize access to foundational medical AI models, and help researchers and medical AI developers rapidly build new solutions. To this end, we open-sourced REMEDIS code-base and we are currently hosting REMEDIS models for chest x-ray and pathology. We expect to add more models and resources for training medical foundation models such as datasets and benchmarks in the future. We also welcome the medical AI research community to contribute to this.
","license":{"name":"Medical AI Foundations Checkpoints: Additional Terms of Service"},"dua":{"name":"Medical AI Foundations Checkpoints: Additional Terms of Service"},"main_storage_size":4134373703,"compressed_storage_size":3583266605,"resource_type":"Model","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/medical-ai-research-foundation/1.0.0/"},{"public_project_uuid":"f7f47719-bf6b-4fa5-b5cf-3b544dc2df71","slug":"mit-critical-datathon-2023","version":"1.0.0","core_doi":"10.13026/8vsr-1f85","version_doi":"10.13026/jfpc-pz79","is_latest_version":true,"short_description":"A SaO2-SpO2 Pairs Dataset derived from MIMIC-IV","publish_date":"2023-05-08","title":"MIT Critical Datathon 2023: a MIMIC-IV Derived Dataset for Pulse Oximetry Correction Models","abstract":"Pulse oximeters are medical devices used to assess peripheral arterial oxygen saturation () noninvasively. In contrast, the "gold standard" requires arterial blood to be drawn to measure the arterial oxygen saturation (). Pulse oximeters currently on the market measure in populations with darker skin tones with lower accuracy. Pulse oximetry inaccuracies can fail to detect episodes of hidden hypoxemia, i.e., low with high . Hidden hypoxemias can result in less treatment and increased mortality. Yet flawed, pulse oximeters remain ubiquitously used because of their ease of use; debiasing the underlying algorithms could alleviate the downstream repercussions of hidden hypoxemia. The dataset supports the building of pulse oximetry correction models. Derived from MIMIC-IV v2.2, it includes 81,797 Intensive Care Unit (ICU) pairs.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":52488339,"compressed_storage_size":13369819,"resource_type":"Challenge","access_policy":"Credentialed","topics":["pulse oximetry","health equity","machine learning"],"source_url":"https://physionet.org/content/mit-critical-datathon-2023/1.0.0/"},{"public_project_uuid":"03cf3141-9cf4-4596-836e-0d784bbc9456","slug":"big-ideas-glycemic-wearable","version":"1.1.1","core_doi":"10.13026/w591-tp72","version_doi":"10.13026/73s9-cw03","is_latest_version":false,"short_description":"Glucose measurements and wrist-worn wearable sensor data from highnormoglycemic participants.","publish_date":"2023-05-11","title":"BIG IDEAs Lab Glycemic Variability and Wearable Device Data","abstract":"This study aimed to determine the feasibility and effectiveness of wearable devices in detecting early physiological changes prior to the development of prediabetes [1-3]. The study generated digital biomarkers for remote, mHealth-based prediabetes and hyperglycemia risk to classify which individuals should undergo further clinical testing. The primary inclusion criteria were subjects aged 35-65 years, inclusive, including only post-menopausal females, with a point of care A1C measurement between 5.2-6.4%, inclusive. Blood was collected during the study for measurement of glucose, hemoglobin A1C, lipoproteins, and triglycerides. Participants wore a Dexcom 6 continuous glucose monitor (CGM) and an Empatica E4 wristband for 10 days while receiving a standardized breakfast meal every other day. At the end of the 10 days, the participant returned to the clinic for an oral glucose tolerance test (OGTT). Research data collected includes physiological measurements from wearable devices such as heart rate, accelerometry, and electrodermal conductance.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":36649624793,"compressed_storage_size":5016598056,"resource_type":"Database","access_policy":"Open","topics":["biomarkers","biomedical engineering","pre-diabetes"],"source_url":"https://physionet.org/content/big-ideas-glycemic-wearable/1.1.1/"},{"public_project_uuid":"476ab607-1c38-4af4-8b4a-d7c72c639ebe","slug":"face-oral-temp-data","version":"1.0.0","core_doi":"10.13026/9ay4-2c37","version_doi":"10.13026/3bhc-9065","is_latest_version":true,"short_description":"Data for each subject include temperatures measured at 29 facial locations over four rounds with two IRTs, oral temperatures measured with a thermometer in two modes, subject demographics (gender, age, ethnicity), environmental conditions, etc.","publish_date":"2023-05-25","title":"Facial and oral temperature data from a large set of human subject volunteers","abstract":"To evaluate the benefit of implementing standardized deployment and acquisition practices in the measurement of elevated body temperature (EBT) with infrared thermographs (IRTs), we conducted a clinical study with more than one thousand subjects. Subject oral temperatures were measured and facial thermal images captured with two evaluated IRTs. Based on the thermal images, temperatures from different locations on the face were extracted based on developed method and are listed in six CSV file as the open database. All data in these files has been de-identified. Based on these data, we published two main clinical study papers, with a focus on IRT effectiveness for EBT measurements and metrics for evaluating IRT clinical accuracy, respectively. Further analysis of this database can still be valuable.
","license":{"name":"Creative Commons Zero 1.0 Universal Public Domain Dedication"},"dua":null,"main_storage_size":4717987,"compressed_storage_size":2516827,"resource_type":"Database","access_policy":"Open","topics":["infectious disease epidemics","thermography","fever screening","thermometry","inner canthus","elevated body temperature","clinical accuracy","receiver operating characteristic curve","facial maximum temperatures","infrared thermograph","pearson correlation coefficients"],"source_url":"https://physionet.org/content/face-oral-temp-data/1.0.0/"},{"public_project_uuid":"37713be4-d30e-40be-a89f-bd4ebf813a8f","slug":"patient-level-data-covid-ms","version":"1.0.0","core_doi":"10.13026/3cmh-xa68","version_doi":"10.13026/feem-fn23","is_latest_version":false,"short_description":"This dataset is part of the Global Data Sharing Initiative. The data was acquired by people with MS and clinicians using a fast data entry tool. The dataset includes demographics, comorbidities and hospital stay and COVID-19 symptoms of PwMS.","publish_date":"2023-05-30","title":"Patient-level dataset to study the effect of COVID-19 in people with Multiple Sclerosis","abstract":"Multiple Sclerosis (MS) is an inflammatory autoimmune disease of the central nervous system, causing increased vulnerability to infections and disability among young adults. Ever since the coronavirus disease 2019 (COVID-19) outbreak, caused by severe acute respiratory syndrome coronavirus 2 infections, there have been concerns among people with MS (PwMS) about the potential interactions between various disease-modifying therapies and COVID-19. The COVID-19 in MS Global Data Sharing Initiative (GDSI) was initiated in 2020 to address these concerns. This paper focuses on the anonymisation and open-sourcing of a GDSI sub-dataset, comprising data entered by people with MS and clinicians using a fast data entry tool. The dataset includes demographics, comorbidities, hospital stay, and COVID-19 symptoms of PwMS. The dataset can be used to perform different statistical analyses to improve our understanding of COVID-19 in MS. Furthermore, this dataset can also be used within the context of educational activities to educate different stakeholders on the complex data science topics that were used within the GDSI.
","license":{"name":"Creative Commons Attribution-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":244512,"compressed_storage_size":24197,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/patient-level-data-covid-ms/1.0.0/"},{"public_project_uuid":"db6d465e-ca01-4c8b-8aba-d08e01b13276","slug":"kinecal","version":"1.0.3","core_doi":"10.13026/8bsb-ch59","version_doi":"10.13026/vvkp-ct80","is_latest_version":true,"short_description":"A dataset for balance falls-risk assessment and balance impairment analysis","publish_date":"2023-06-08","title":"KINECAL","abstract":"The field of human action recognition has made great strides in recent years, much helped by the availability of a wide variety of datasets that use Kinect to record human movement. Conversely, progress towards using Kinect in clinical practice has been hampered by the lack of appropriate data. In particular, datasets that contain clinically significant movements and appropriate metadata. This paper proposes a dataset to address this issue, namely KINECAL. It contains the recordings of 90 individuals carrying out 11 movements, commonly used in the clinical assessment of balance. The dataset contains relevant metadata, including clinical labelling, falls history and postural sway metrics. In addition, the recordings were made in both the lab and in informal settings.
","license":{"name":"Creative Commons Zero 1.0 Universal Public Domain Dedication"},"dua":null,"main_storage_size":193424110739,"compressed_storage_size":92516089893,"resource_type":"Database","access_policy":"Open","topics":["balance","posturography","clinical tests","postural sway","falls-risk","age-related changes"],"source_url":"https://physionet.org/content/kinecal/1.0.3/"},{"public_project_uuid":"41c8c442-fbe1-423f-92b4-dd6fdf4f157a","slug":"sicdb","version":"1.0.6","core_doi":"10.13026/ezs8-6v88","version_doi":"10.13026/dm9f-dm56","is_latest_version":false,"short_description":"The SICdb dataset, version 1.0.6 contains 27386 admissions to an ICU in an Austrian tertiary care institution.","publish_date":"2023-06-12","title":"Salzburg Intensive Care database (SICdb), a freely accessible intensive care database","abstract":"The SICdb dataset offers insights into over 27 thousand intensive care admissions, including therapies and data on preceding surgeries. Data were collected between 2013 and 2021 from four different intensive care units at the University Hospital Salzburg, having more than 3 thousand intensive care admissions per year on 41 beds. The dataset is deidentified and contains, amongst others, case information, vital signs, laboratory results and medication data. SICdb provides both aggregated once-per-hour and highly granular once-per-minute data, making it suitable for computational and machine learning-based research.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":2372225408,"compressed_storage_size":2369632252,"resource_type":"Database","access_policy":"Contributor Review","topics":["intensive care","critical care","clinical","machine learning","open data"],"source_url":"https://physionet.org/content/sicdb/1.0.6/"},{"public_project_uuid":"47e1017c-8c0e-40ad-b4df-5231bbb56922","slug":"cxr-lt-iccv-workshop-cvamd","version":"1.0.0","core_doi":"10.13026/5mbq-9k80","version_doi":"10.13026/721s-vs37","is_latest_version":false,"short_description":"In this challenge, participants will develop methods for long-tailed, multi-label thorax disease classification on chest X-rays. CXR-LT is held in conjunction with the ICCV 2023 workshop, Computer Vision for Automated Medical Diagnosis (CVAMD).","publish_date":"2023-06-12","title":"CXR-LT: Multi-Label Long-Tailed Classification on Chest X-Rays","abstract":"Many real-world problems, including diagnostic medical imaging exams, are “long-tailed” – there are a few common findings followed by more relatively rare conditions. In chest radiography, diagnosis is both a long-tailed and multi-label problem, as patients often present with multiple disease findings simultaneously. This is distinct from most large-scale image classification benchmarks, where each image only belongs to one label and the distribution of labels is relatively balanced. While researchers have begun to study the problem of long-tailed learning in medical image recognition, few have studied its interplay with label co-occurrence. This competition will provide a challenging large-scale multi-label long-tailed learning task on chest X-rays (CXRs), encouraging community engagement with this emerging interdisciplinary topic.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":74721996,"compressed_storage_size":18827760,"resource_type":"Challenge","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/cxr-lt-iccv-workshop-cvamd/1.0.0/"},{"public_project_uuid":"cd4d47ff-27bf-403b-9913-6bff7034d1d9","slug":"mimicel-ed","version":"2.1.0","core_doi":"10.13026/j8yc-gg94","version_doi":"10.13026/c9yj-1t90","is_latest_version":true,"short_description":"MIMIC-IV Event Log for Emergency Department","publish_date":"2023-06-16","title":"MIMICEL: MIMIC-IV Event Log for Emergency Department","abstract":"In this work, we extract an event log from the MIMIC-IV-ED dataset by adopting a well-established event log generation methodology, and we name this event log MIMICEL. The data tables in the MIMIC-IV-ED dataset relate to each other based on the existing relational database schema, and each table records the individual activities of patients along their journey in the emergency department (ED). While the data tables in the MIMIC-IV-ED dataset catch snapshots of a patient journey in the ED, the extracted event log MIMICEL aims to capture an end-to-end process of the patient journey. This will enable us to analyse the existing patient flows, thereby improving the efficiency of an ED process.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":11602638592,"compressed_storage_size":567232288,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimicel-ed/2.1.0/"},{"public_project_uuid":"2afd116c-dc28-4ec5-b6f9-6aa99e0b13d3","slug":"i-care","version":"2.0","core_doi":"10.13026/sd64-2856","version_doi":"10.13026/avek-0p97","is_latest_version":false,"short_description":"The clinical and EEG data for this dataset originates from seven academic hospitals in the U.S. and Europe led by investigators part of the International Cardiac Arrest REsearch consortium (I-CARE).","publish_date":"2023-06-16","title":"I-CARE: International Cardiac Arrest REsearch consortium Database","abstract":"The International Cardiac Arrest REsearch consortium (I-CARE) Database includes baseline clinical information and continuous electroencephalogram (EEG) and electrocardiogram (ECG) recordings from comatose patients following cardiac arrest. The patients were admitted to an intensive care unit (ICU) in one of seven academic hospitals in the U.S. and Europe and monitored for several hours to several days. The long-term neurological function of the patients was determined using the Cerebral Performance Category scale.
","license":{"name":"Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":1693971891026,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/i-care/2.0/"},{"public_project_uuid":"65766fae-54f1-46ca-9213-dcd33a2c35c5","slug":"rexval-dataset","version":"1.0.0","core_doi":"10.13026/r7wq-9h91","version_doi":"10.13026/2fp8-qr71","is_latest_version":true,"short_description":"The Radiology Report Expert Evaluation (ReXVal) Dataset is a publicly available dataset of radiologist evaluations of errors in automatically generated radiology reports.","publish_date":"2023-06-20","title":"Radiology Report Expert Evaluation (ReXVal) Dataset","abstract":"The Radiology Report Expert Evaluation (ReXVal) Dataset is a publicly available dataset of radiologist evaluations of errors in automatically generated radiology reports. The dataset contains annotations from 6 board certified radiologists on clinically significant and clinically insignificant errors under 6 error categories for candidate radiology reports with respect to ground-truth reports from the MIMIC-CXR dataset. There are 4 candidate reports generated for 50 studies, translating to 200 pairs of candidate and ground-truth reports on which radiologists provided annotations. The dataset has been used to evaluate the alignment between scoring of automated metrics and that of radiologists, investigate the failure modes of automated metrics, and build a composite automated metric, in a study on how to meaningfully measure progress in radiology report generation. It is also created to support additional medical AI research in radiology and other expert tasks.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":456133,"compressed_storage_size":107347,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/rexval-dataset/1.0.0/"},{"public_project_uuid":"1e40596e-5b0e-4929-8807-97fc680faaa5","slug":"chexmask-cxr-segmentation-data","version":"0.1","core_doi":"10.13026/6eky-y831","version_doi":"10.13026/dx54-8351","is_latest_version":false,"short_description":"CheXmask Database is a 676,803 uniformly annotated chest radiographs with segmentation masks. Images were segmented using HybridGNet, with automatic quality control indicated by RCA scores.","publish_date":"2023-06-27","title":"CheXmask Database: a large-scale dataset of anatomical segmentation masks for chest x-ray images","abstract":"The CheXmask Database presents a comprehensive, uniformly annotated collection of chest radiographs, constructed from six public databases: CANDID-PTX, ChestX-ray8, Chexpert, MIMIC-CXR-JPG, Padchest and VinDr-CXR. The database aggregates 676,803 anatomical segmentation masks derived from images which have been processed using the HybridGNet model to ensure consistent, high-quality segmentation. To confirm the quality of the segmentations, we include in this database individual Reverse Classification Accuracy (RCA) scores for each of the segmentation masks. This dataset is intended to catalyze further innovation and refinement in the field of semantic chest X-ray analysis, offering a significant resource for researchers in the medical imaging domain.
","license":{"name":"Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":40457161699,"compressed_storage_size":16942088879,"resource_type":"Database","access_policy":"Open","topics":["medical image segmentation","automatic quality assesment","chest x-ray segmentation"],"source_url":"https://physionet.org/content/chexmask-cxr-segmentation-data/0.1/"},{"public_project_uuid":"b08374f2-cfda-44fb-8eb7-ccb0b4cb4160","slug":"ptb-xl-plus","version":"1.0.1","core_doi":"10.13026/nqsf-pc74","version_doi":"10.13026/g6h6-7g88","is_latest_version":true,"short_description":"ECG feature dataset accompanying the PTB-XL ECG dataset","publish_date":"2023-06-27","title":"PTB-XL+, a comprehensive electrocardiographic feature dataset","abstract":"The PTB-XL+ dataset is a comprehensive feature dataset that supplements the PTB-XL ECG dataset. It includes ECG features extracted via two commercial and one open-source algorithm in a harmonized format as well as median beats or, where applicable, fiducial points extracted by the respective algorithms. In addition, it provides automatic diagnosis statements from one commercial ECG analysis algorithm at different processing levels that are ready to be used for training and evaluation of machine learning models.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":2182061968,"compressed_storage_size":733758236,"resource_type":"Database","access_policy":"Open","topics":["ecg","electrocardiography","ptb-xl","ptb"],"source_url":"https://physionet.org/content/ptb-xl-plus/1.0.1/"},{"public_project_uuid":"f4800eb7-3c9b-42d3-ac97-5f49a95e1b28","slug":"discharge-summary-templates","version":"1.0","core_doi":"10.13026/ngc0-0f54","version_doi":"10.13026/63ab-qf77","is_latest_version":true,"short_description":"Annotated MIMIC-IV discharge summaries used to explore deidentification of names","publish_date":"2023-07-05","title":"Annotated MIMIC-IV discharge summaries for a study on deidentification of names","abstract":"We introduce a dataset of people’s names and clinical note templates to support research on the demographic bias of de-identification systems. The dataset contains 16 name sets that vary along four demographic dimensions: gender, race, name popularity, and the decade of popularity. It also includes 100 clinical templates manually curated from MIMIC-IV discharge summaries. These templates are then populated with names generated from the 16 name sets to investigate the demographic bias in nine public and private de-identification methods. The dataset provides a valuable resource for researchers who are interested in understanding and mitigating the unfairness of de-identification systems.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":416611683,"compressed_storage_size":31610896,"resource_type":"Database","access_policy":"Credentialed","topics":["fairness","deidentification"],"source_url":"https://physionet.org/content/discharge-summary-templates/1.0/"},{"public_project_uuid":"a7bb4e9b-04a6-4570-9bc4-7b230ef9ea56","slug":"corpus-fungal-infections","version":"1.0.0","core_doi":"10.13026/g3t3-2c54","version_doi":"10.13026/fmj9-p237","is_latest_version":false,"short_description":"A corpus of cytology and histopathology reports annotated for terminology relevant to fungal infections. Ideal for validation of named entity recognition and relation extraction methods.","publish_date":"2023-07-12","title":"CHIFIR: Cytology and Histopathology Invasive Fungal Infection Reports","abstract":"Surveillance of invasive fungal infection (IFI) in clinical settings is a laborious process requiring a detailed review of patient medical history. One of the key sources of clinical information is cytology and histopathology reports: pathologist-produced free-text reports outlining the macroscopic and microscopic structure of a specimen. The data was generated to facilitate the development of an automated tool for the detection of IFI. The Cytology and Histopathology Invasive Fungal Infection Reports (CHIFIR) corpus contains 283 de-identified reports annotated by infectious diseases physicians for terminology relevant to the IFI diagnosis. These include IFI-specific concepts and certainty cues, as well as relations between them. We release the annotation schema and the original reports along with the corresponding annotation files. We anticipate the CHIFIR corpus to be useful in the development and validation of named entity recognition and relation extraction methods with a focus on clinical data. Such methods can be instrumental in processing other types of clinical documentation (radiology reports, clinical notes, nursing notes) with various downstream tasks in mind.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":495508,"compressed_storage_size":399456,"resource_type":"Database","access_policy":"Credentialed","topics":["invasive fungal infections","nlp","information extraction","clinical documentation"],"source_url":"https://physionet.org/content/corpus-fungal-infections/1.0.0/"},{"public_project_uuid":"aeb573b8-9a44-4a5c-9f21-8c24c35d36dc","slug":"eeg-eye-gaze-data","version":"1.0.0","core_doi":"10.13026/9m3f-ac20","version_doi":"10.13026/qj5m-n649","is_latest_version":true,"short_description":"The brain activity and eye gaze data were recorded from 25 participants performing surgical tasks using a robot simulator. The performance score was created by the simulator. Data can be used to evaluate surgical performance.","publish_date":"2023-07-14","title":"Electroencephalogram and eye-gaze datasets for robot-assisted surgery performance evaluation","abstract":"Surgical training and patient safety depend on objective and accurate performance evaluation. Robot-assisted surgery (RAS) is newer and its training less well defined, but resources are not easily accessed to support data collection from participants with various RAS skill levels. This work describes and makes available for public use electroencephalogram (EEG) and eye-gaze data collected from 25 participants performing RAS simulator tasks with varying levels of RAS experience.Twenty-seven tasks from six modules of the da Vinci®;Skills Simulator ™, which are standard elements of most RAS surgical skills training curricula, were taken into consideration. The dataset contains performance scores (scale: 0 to 100) generated by the simulator after each exercise was completed. Each task was completed by the participants at least twice, and if they weren't able to complete it with a score of 70 (out of 100) on at least one try, they tried again until they did. The National Institute of Biomedical Imaging and Bioengineering (NIBIB) and Roswell Park Comprehensive Cancer Center (RPCCC) supported the collection of data for use in RAS training, which is why the dataset has been named the NIBIB-RPCCC-RAS performance evaluation dataset. The dataset contains 1636 EEG recordings (in the "*.edf" format), 1559 eye-gaze recordings (in the "*.csv" format), performance scores, and participant demographic data, such as age, gender, and dominant hand. To our knowledge, this is the first shared dataset that contains EEG and eye-gaze data from participants with a variety of RAS experiences who performed all tasks included in the da Vinci simulator modules.
","license":{"name":"Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":35617115460,"compressed_storage_size":9897835644,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/eeg-eye-gaze-data/1.0.0/"},{"public_project_uuid":"3f2881ec-1a49-48a8-a03f-4797fad4121d","slug":"eeg-gaba-anesthesia","version":"1.0.0","core_doi":"10.13026/7ygq-3f91","version_doi":"10.13026/dx44-kw30","is_latest_version":true,"short_description":"Electroencephalogram dynamics during unconsciousness mediated by GABAergic anesthetics.","publish_date":"2023-07-21","title":"Electroencephalogram dynamics during unconsciousness mediated by GABAergic-anesthetics","abstract":"This dataset of electroencephalogram (EEG) activity was collected from subjects undergoing general anesthesia using GABAergic agents. The typical EEG of subjects during unconsciousness mediated by GABAergic anesthetics shows slow-delta oscillations (0.3 - 4Hz) and alpha oscillations (8 – 14Hz). At higher doses, these oscillations devolve into burst suppression, a state of profound brain inactivation. This data was used to investigate the transition between these states of unconsciousness. It was collected as part of a larger study to investigate the changes in brain activity during unconsciousness mediated by GABAergic anesthetics. All data collection and experimental protocols were approved by the Mass General Brigham Human Research Committee.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":69454237,"compressed_storage_size":19729092,"resource_type":"Database","access_policy":"Contributor Review","topics":[],"source_url":"https://physionet.org/content/eeg-gaba-anesthesia/1.0.0/"},{"public_project_uuid":"4bcff4ee-e8ed-43aa-bde4-ff54516aa426","slug":"mimic-iv-ecg","version":"0.3","core_doi":"10.13026/b95v-ff39","version_doi":"10.13026/dp3j-2c96","is_latest_version":false,"short_description":"The MIMIC-IV ECG module contains approximately 800,000 diagnostic electrocardiograms across nearly 160,000 unique patients. These patients overlap with the patients from the MIMIC-IV Clinical Database.","publish_date":"2023-07-21","title":"MIMIC-IV-ECG - Diagnostic Electrocardiogram Matched Subset","abstract":"The MIMIC-IV-ECG module contains approximately 800,000 diagnostic electrocardiograms across nearly 160,000 unique patients. These diagnostic ECGs use 12 leads and are 10 seconds in length. They are sampled at 500 Hz. This subset contains all of the ECGs for patients who appear in the MIMIC-IV Clinical Database. When a cardiologist report is available for a given ECG, it is also provided. The patients in MIMIC-IV-ECG have been matched against the MIMIC-IV Clinical Database, making it possible to link to information across the MIMIC-IV modules.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":97094584079,"compressed_storage_size":36304409864,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimic-iv-ecg/0.3/"},{"public_project_uuid":"4dd14919-1435-40e9-97ea-3cae45df0a21","slug":"mimic-iv-echo","version":"0.1","core_doi":"10.13026/agyq-zp57","version_doi":"10.13026/ef48-v217","is_latest_version":false,"short_description":"The MIMIC-IV-ECHO module contains more than 500,000 echocardiograms across more than 4,500 unique patients. These patients overlap with the patients from the MIMIC-IV Clinical Database.","publish_date":"2023-07-21","title":"MIMIC-IV-ECHO: Echocardiogram Matched Subset","abstract":"The MIMIC-IV-ECHO module contains more than 500,000 echocardiograms across 7,243 studies from 4,579 distinct patients. A given study consists of numerous sequences of images, with each sequence representing a particular view of the patient's heart. This subset contains echocardiograms for patients who appear in the MIMIC-IV Clinical Database and were admitted between 2017 and 2019. We have provided information for linking a given echocardiogram to the cardiologist's report where available. Records in MIMIC-IV-ECHO are matched to the MIMIC-IV Clinical Database, making it possible to link to information across the MIMIC-IV modules..
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1896550850641,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimic-iv-echo/0.1/"},{"public_project_uuid":"152b6bfb-1efa-479a-8935-eea12f791a4a","slug":"orchid","version":"1.0.0","core_doi":"10.13026/b1c0-3506","version_doi":"10.13026/eytj-4f29","is_latest_version":false,"short_description":"Multi-center dataset on organ procurement in the United States","publish_date":"2023-07-25","title":"Organ Retrieval and Collection of Health Information for Donation (ORCHID)","abstract":"There are well-documented inefficiencies and inequities in the current system of deceased donor organ transplantation. While much prior research has focused on designing better allocation systems to distribute donated organs, more can be done to study organ procurement, the process by which organs are recovered from deceased donors. Improvements to organ procurement can greatly increase the supply of a scarce public resource, improving health outcomes for the more than 100,000 patients currently waiting for an organ transplant. Such changes will also impact health equity, as a disproportionate number of waitlisted patients come from marginalized communities. So far, research into organ procurement has been hamstrung by a lack of transparent data. To address this pressing issue, we have collected the first ever multi-center dataset on organ procurement, covering ten years of clinical, financial, and administrative information from six organ procurement organizations (OPOs). This novel dataset can be used to identify opportunities to improve current practice, build tools to help OPOs make better decisions, and recommend policy changes to increase procurement.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":28168469,"compressed_storage_size":3971421,"resource_type":"Database","access_policy":"Restricted","topics":["organ transplantation","organ procurement organizations"],"source_url":"https://physionet.org/content/orchid/1.0.0/"},{"public_project_uuid":"486bfafd-183b-4bbd-a71a-c33e6102bae2","slug":"heart-lung-segmentations-data","version":"1.0.0","core_doi":"10.13026/vw06-hz66","version_doi":"10.13026/0k35-mb65","is_latest_version":true,"short_description":"Heart and lung segmentations for 200 MIMIC-CXR/MIMIC-CXR-JPG chest x-rays and heart segmentations for 138 Montgomery County tuberculosis chest X-rays.","publish_date":"2023-08-14","title":"Heart and lung segmentations for MIMIC-CXR/MIMIC-CXR-JPG and Montgomery County TB databases","abstract":"Segmenting the heart and lungs from chest X-ray images is essential for accurate disease diagnosis, enabling the calculation of image-derived digital biomarkers for cardiopulmonary health assessment. Furthermore, creation of annotated data sets can enhance development of machine learning applications in disease detection, reduce image noise for better clarity, or promote standardization for tracking disease progression or treatment effects. Generation of manual segmentations is time intensive process and it also requires access to trained medical professionals to verify the quality and accuracy of the scans. Originally made to train heart and lung segmentation/detection models for cardiomegaly diagnosis, the manual segmentations in this data paper are published here in hope of aiding development of AI models relating to heart and lungs identification in chest X-rays. This database presents the heart and lung segmentations for 200 semi-randomly chosen MIMIC-CXR/MIMIC-CXR-JPG posterior-anterior chest X-rays for the purpose of training detection and segmentation networks. Additionally, it contains the heart segmentations for the 138 posterior-anterior chest X-rays in the Montgomery Country tuberculosis database.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":7371779,"compressed_storage_size":3103294,"resource_type":"Database","access_policy":"Open","topics":["heart and lungs","montgomery country tb","segmentation","mimic-cxr"],"source_url":"https://physionet.org/content/heart-lung-segmentations-data/1.0.0/"},{"public_project_uuid":"9015186f-0786-4eed-93ed-a4c238623c05","slug":"refisco","version":"1.0","core_doi":"10.13026/0neq-hp96","version_doi":"10.13026/cneg-zk64","is_latest_version":true,"short_description":"Preliminary human expert evaluation study on 60 MIMIC-CXR radiology reports","publish_date":"2023-08-23","title":"ReFiSco: Report Fix and Score Dataset for Radiology Report Generation","abstract":"Automated generation of clinically accurate radiology reports can improve patient care. In order to improve automatic report generation, it is helpful to understand what types of errors are common in generated reports. Thus, we introduce the Report Fix and Score Dataset for Radiology Reports (ReFiSco-v0), which was collected through an institutional review board-approved study. In our study, we recruit radiologists to provide expert evaluations on a subset of 60 studies from MIMIC-CXR. For each radiology image, we compile three reports: one generated from the model X-REM, one from the model CXR-RePaiR trained on the same MIMIC-CXR training set, and one taken from a human benchmark (MIMIC-CXR). To each radiologist, we present one image and one report for each of the 60 studies. Each report is randomly and independently chosen from one of the three sources. The radiologist is blinded to the source. We ask each radiologist to assess the error severity of their assigned reports.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":274037,"compressed_storage_size":79993,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/refisco/1.0/"},{"public_project_uuid":"536ab55a-58af-4078-80fc-9dc597db42a8","slug":"eeg-eye-gaze-for-fls-tasks","version":"1.0.0","core_doi":"10.13026/c8sa-ed22","version_doi":"10.13026/kyjw-p786","is_latest_version":true,"short_description":"Brain activity and eye gaze data were collected from a group of 25 participants who completed the FLS tasks using a trainer box (Pyxus®). Each participant performed the tasks five times, and their performance was evaluated by an expert rater.","publish_date":"2023-08-23","title":"Integration of Electroencephalogram and Eye-Gaze Datasets for Performance Evaluation in Fundamentals of Laparoscopic Surgery (FLS) Tasks","abstract":"Laparoscopic surgery represents a significant advancement in healthcare, aiming to enhance patient outcomes. However, evaluating performance in laparoscopic surgery tasks can be challenging due to subjective assessments. To address this, this study provides a comprehensive description and public access to electroencephalogram (EEG) and eye-gaze data collected from 25 participants with varying levels of robotic surgery experience. The participants performed Fundamentals of Laparoscopic Surgery (FLS) tasks using a trainer box (Pyxus®), focusing on three standard FLS elements. The dataset, named the NIBIB-RPCCC-FLS dataset, includes performance scores assessed by a rater after each task attempt. Each participant completed the tasks five times, with some subjects performing them only twice. The data collection received support from the National Institute of Biomedical Imaging and Bioengineering (NIBIB) and Roswell Park Comprehensive Cancer Center (RPCCC), with the goal of facilitating objective performance evaluation in FLS tasks. The dataset consists of 315 EEG recordings in the ".edf" format, 315 eye-gaze recordings in the ".csv" format, performance scores, and participant demographic data, such as age, gender, dominant hand, and dominant eye. Notably, this dataset is the first shared collection to include EEG and eye-gaze data from participants with different experiences in robotic surgery who performed FLS tasks.
","license":{"name":"Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":8330833087,"compressed_storage_size":2949084706,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/eeg-eye-gaze-for-fls-tasks/1.0.0/"},{"public_project_uuid":"76537f5c-7a55-490a-8d7b-ac564f8998d5","slug":"multimodal-dental-dataset","version":"1.0.0","core_doi":"10.13026/bxzm-dg78","version_doi":"10.13026/s5z3-2766","is_latest_version":false,"short_description":"A new dental dataset that contains 389 patients, three commonly used dental image models, and images of various health conditions of the oral cavity.","publish_date":"2023-09-06","title":"A multimodal dental dataset facilitating machine learning research and clinic services","abstract":"Oral diseases affect nearly 3.5 billion people, with the majority residing in low- and middle-income countries. Due to limited healthcare resources, many individuals are unable to access proper oral healthcare services. Image-based machine learning technology is one of the most promising approaches to improving oral healthcare services and reducing patient costs. Openly accessible datasets play a crucial role in facilitating the development of machine learning techniques. However, existing dental datasets have limitations such as a scarcity of Cone Beam Computed Tomography (CBCT) data, lack of matched multi-modal data, and insufficient complexity and diversity of the data. This project addresses these challenges by providing a dataset that includes 574 CBCT images from 389 patients, multi-modal data with matching modalities, and images representing various oral health conditions.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":197781413537,"compressed_storage_size":88617533423,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/multimodal-dental-dataset/1.0.0/"},{"public_project_uuid":"958a00e7-dc6e-4448-95a0-2f97d1806467","slug":"chexmask-cxr-segmentation-data","version":"0.2","core_doi":"10.13026/6eky-y831","version_doi":"10.13026/fkmr-dz62","is_latest_version":false,"short_description":"CheXmask Database is a 676,803 uniformly annotated chest radiographs with segmentation masks. Images were segmented using HybridGNet, with automatic quality control indicated by RCA scores.","publish_date":"2023-09-06","title":"CheXmask Database: a large-scale dataset of anatomical segmentation masks for chest x-ray images","abstract":"The CheXmask Database presents a comprehensive, uniformly annotated collection of chest radiographs, constructed from six public databases: CANDID-PTX, ChestX-ray8, Chexpert, MIMIC-CXR-JPG, Padchest and VinDr-CXR. The database aggregates 676,803 anatomical segmentation masks derived from images which have been processed using the HybridGNet model to ensure consistent, high-quality segmentation. To confirm the quality of the segmentations, we include in this database individual Reverse Classification Accuracy (RCA) scores for each of the segmentation masks. This dataset is intended to catalyze further innovation and refinement in the field of semantic chest X-ray analysis, offering a significant resource for researchers in the medical imaging domain.
","license":{"name":"Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":40457161699,"compressed_storage_size":16942088879,"resource_type":"Database","access_policy":"Open","topics":["chest x-ray segmentation","medical image segmentation","automatic quality assesment"],"source_url":"https://physionet.org/content/chexmask-cxr-segmentation-data/0.2/"},{"public_project_uuid":"b80a4db8-3fd0-48e7-ae2a-8a3a9932587f","slug":"mimic-iv-ecg","version":"1.0","core_doi":"10.13026/b95v-ff39","version_doi":"10.13026/4nqg-sb35","is_latest_version":true,"short_description":"The MIMIC-IV ECG module contains approximately 800,000 diagnostic electrocardiograms across nearly 160,000 unique patients. These patients overlap with the patients from the MIMIC-IV Clinical Database.","publish_date":"2023-09-15","title":"MIMIC-IV-ECG: Diagnostic Electrocardiogram Matched Subset","abstract":"The MIMIC-IV-ECG module contains approximately 800,000 diagnostic electrocardiograms across nearly 160,000 unique patients. These diagnostic ECGs use 12 leads and are 10 seconds in length. They are sampled at 500 Hz. This subset contains all of the ECGs for patients who appear in the MIMIC-IV Clinical Database. When a cardiologist report is available for a given ECG, we provide the needed information to link the waveform to the report. The patients in MIMIC-IV-ECG have been matched against the MIMIC-IV Clinical Database, making it possible to link to information across the MIMIC-IV modules.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":97017195666,"compressed_storage_size":36295288053,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/mimic-iv-ecg/1.0/"},{"public_project_uuid":"9b7a28ef-057e-4e3a-9a7f-15cbe1d964e5","slug":"medical-diff-vqa","version":"1.0.0","core_doi":"10.13026/hbsg-tk87","version_doi":"10.13026/5jes-bx23","is_latest_version":false,"short_description":"MIMIC-Diff-VQA provides a large-scale dataset for Difference visual question answering in medical chest x-ray images.","publish_date":"2023-09-15","title":"Medical-Diff-VQA: A Large-Scale Medical Dataset for Difference Visual Question Answering on Chest X-Ray Images","abstract":"The task of Difference Visual Question Answering involves answering questions about the difference between a pair of main and reference images. This process is consistent with the radiologist's diagnosis practice that compares the current image with the reference before concluding the report. We've assembled a new dataset, called Medical-Diff-VQA, for this purpose. Unlike previous medical VQA datasets, ours is the first one designed specifically for the Difference Visual Question Answering task, with questions crafted to suit the Assessment-Diagnosis-Intervention-Evaluation treatment procedure employed by medical professionals. The Medical-Diff-VQA dataset, a derivative of the MIMIC-CXR dataset, consists of questions categorized into seven categories: abnormality (145,421), location (84,193), type (27,478), level (67,296), view (56,265), presence (155,726), and difference(164,324). The 'difference' questions are specifically for comparing two images. In total, the Medical-Diff-VQA dataset contains 700,703 question-answer pairs derived from 164,324 pairs of main and reference images.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":312419307,"compressed_storage_size":21843987,"resource_type":"Database","access_policy":"Credentialed","topics":["difference vqa","chest x-ray","visual question answering","vqa","difference visual question answering"],"source_url":"https://physionet.org/content/medical-diff-vqa/1.0.0/"},{"public_project_uuid":"6043903b-0ef0-43df-8057-6f40f725d11d","slug":"big-ideas-glycemic-wearable","version":"1.1.2","core_doi":"10.13026/w591-tp72","version_doi":"10.13026/zthx-5212","is_latest_version":false,"short_description":"Glucose measurements and wrist-worn wearable sensor data from highnormoglycemic participants.","publish_date":"2023-09-18","title":"BIG IDEAs Lab Glycemic Variability and Wearable Device Data","abstract":"This study aimed to determine the feasibility and effectiveness of wearable devices in detecting early physiological changes prior to the development of prediabetes [1-3]. The study generated digital biomarkers for remote, mHealth-based prediabetes and hyperglycemia risk to classify which individuals should undergo further clinical testing. The primary inclusion criteria were subjects aged 35-65 years, inclusive, including only post-menopausal females, with a point of care A1C measurement between 5.2-6.4%, inclusive. Blood was collected during the study for measurement of glucose, hemoglobin A1C, lipoproteins, and triglycerides. Participants wore a Dexcom 6 continuous glucose monitor (CGM) and an Empatica E4 wristband for 10 days while receiving a standardized breakfast meal every other day. At the end of the 10 days, the participant returned to the clinic for an oral glucose tolerance test (OGTT). Research data collected includes physiological measurements from wearable devices such as heart rate, accelerometry, and electrodermal conductance.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":36645548198,"compressed_storage_size":5015249922,"resource_type":"Database","access_policy":"Open","topics":["biomedical engineering","pre-diabetes","biomarkers"],"source_url":"https://physionet.org/content/big-ideas-glycemic-wearable/1.1.2/"},{"public_project_uuid":"1330dba5-0a33-46a8-81cd-2e94090c6e7c","slug":"inspire","version":"0.1","core_doi":"10.13026/dv8j-5412","version_doi":"10.13026/mw0m-9a42","is_latest_version":false,"short_description":"A public dataset that contains information related to surgery, anesthesia, laboratory results, medications, diagnosis, and outcomes from 50% of the patients who received surgery at Seoul National University Hospital between 2011 and 2020.","publish_date":"2023-09-19","title":"INSPIRE, a publicly available research dataset for perioperative medicine","abstract":"We present the INSPIRE dataset, a publicly available research dataset in perioperative medicine, which includes approximately 130,000 cases (50% of the surgical cases) who underwent anesthesia for surgery at an academic institution in South Korea over a ten-year period between 2011 and 2020. (For version 0.1, only about 52,000 cases, 20% of the surgical cases, were included.) This comprehensive dataset includes patient characteristics such as age, sex, American Society of Anesthesiologists physical status classification, diagnosis, surgical procedure code, department, and type of anesthesia. It also includes vital signs in the operating theatre, general wards, and intensive care units (ICUs), laboratory results from six months before admission to six months after discharge, and medication during hospitalization. Complications include total hospital and ICU length of stay and in-hospital death. We hope this dataset will inspire collaborative research and development in perioperative medicine and serve as a reproducible external validation dataset to improve surgical outcomes.
","license":{"name":"Korea Credentialed Health Data License-1-0-0"},"dua":{"name":"Korea Credentialed Health Data Agreement-1-0-0"},"main_storage_size":1860292049,"compressed_storage_size":229890636,"resource_type":"Database","access_policy":"Credentialed","topics":["surgery","perioperative medicine","multi-center","open dataset"],"source_url":"https://physionet.org/content/inspire/0.1/"},{"public_project_uuid":"822b7669-2888-48f9-9b09-741e34ed1f86","slug":"icehg-ds","version":"1.0.0","core_doi":"10.13026/vrvd-mf28","version_doi":"10.13026/rfts-nv79","is_latest_version":false,"short_description":"The design and development of ICEHG DS was funded by the Slovenian Research Agency (ARRS) under the research project Metabolic and inborn factors of reproductive health, birth III.","publish_date":"2023-09-22","title":"Induced Cesarean EHG DataSet (ICEHG DS): An open dataset with electrohysterogram records of pregnancies ending in induced and cesarean section delivery","abstract":"The surface electrohysterogram (EHG) has emerged as a promising diagnostic tool for non-invasive automated preterm birth prediction. The existing automated preterm birth prediction methods rely on the use of uterine EHG records coming from spontaneous preterm and term deliveries, and are indifferent to term induced and cesarean deliveries. In order to enhance current publicly available pool of term EHG records, we developed a new EHG dataset containing 126 30-minute EHG records, recorded either early (23rd week), or later during pregnancy (31st week), of those pregnancies that were expected to end in spontaneous term delivery, but ended in induced or cesarean section delivery. The records were collected at the University Medical Center Ljubljana, Ljubljana, Slovenia. The dataset includes 38 and 43, early and later, induced; 11 and 8, early and later, cesarean; and 13 and 13, early and later, induced-cesarean records. This database enables better understanding of the underlying physiological mechanisms involved during pregnancies ending in induced and cesarean deliveries, and provides more robust and realistic assessment of the performance of automated preterm birth prediction.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":121177789,"compressed_storage_size":93277918,"resource_type":"Database","access_policy":"Open","topics":["neuroelectric","pregnancy","induced delivery","electrohysterogram","cesarean-section delivery"],"source_url":"https://physionet.org/content/icehg-ds/1.0.0/"},{"public_project_uuid":"6764c987-33ab-4b2c-95c1-172b4016ce5d","slug":"cxr-lt-iccv-workshop-cvamd","version":"1.1.0","core_doi":"10.13026/5mbq-9k80","version_doi":"10.13026/c4tr-kr83","is_latest_version":false,"short_description":"CXR-LT 2023 was a challenge for long-tailed, multi-label thorax disease classification on chest X-rays, held in conjunction with the ICCV 2023 workshop, CVAMD. This page contains extended long-tailed versions of the MIMIC-CXR-JPG v2.0.0 dataset.","publish_date":"2023-09-28","title":"CXR-LT: Multi-Label Long-Tailed Classification on Chest X-Rays","abstract":"Many real-world problems, including diagnostic medical imaging exams, are “long-tailed” – there are a few common findings followed by more relatively rare conditions. In chest radiography, diagnosis is both a long-tailed and multi-label problem, as patients often present with multiple disease findings simultaneously. This is distinct from most large-scale image classification benchmarks, where each image only belongs to one label and the distribution of labels is relatively balanced. While researchers have begun to study the problem of long-tailed learning in medical image recognition, few have studied its interplay with label co-occurrence. This competition will provide a challenging large-scale multi-label long-tailed learning task on chest X-rays (CXRs), encouraging community engagement with this emerging interdisciplinary topic. This project contains labels for the CXR-LT 2023 competition dataset, containing 377,110 CXRs from 26 classes, and a related subset used in the MICCAI 2023 paper, "How Does Pruning Impact Multi-Label Long-Tailed Learning?" containing 257,018 frontal CXRs from 19 classes.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":178064323,"compressed_storage_size":54058562,"resource_type":"Challenge","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/cxr-lt-iccv-workshop-cvamd/1.1.0/"},{"public_project_uuid":"04012464-96f0-447f-8e1f-ef38f578d2aa","slug":"challenge-2022","version":"1.0.0","core_doi":"10.13026/dxqa-pb41","version_doi":"10.13026/t49p-5v35","is_latest_version":true,"short_description":"2022 Physionet Challenge is devoted to detecting the presence or absence of murmurs from multiple heart sound recordings from multiple auscultation locations, as well as detecting the clinical outcomes.","publish_date":"2023-09-28","title":"Heart Murmur Detection from Phonocardiogram Recordings: The George B. Moody PhysioNet Challenge 2022","abstract":"Congenital heart diseases affect about 1% of newborns, representing an important morbidity and mortality factor for several severe conditions, including advanced heart failure [1]. In a 2019 survey, it was estimated that congenital heart diseases affect over 500,000 children in East Africa [2], and about 8 in every 1000 live births [3]. Acquired heart diseases include rheumatic fever and the Kawasaki disease, the former being a serious public health problem in developing regions, e.g., rural Brazil [4]. Several regions of developing countries have difficulties in diagnosing and treating both congenital and acquired heart conditions in children. This is mainly due to the lack of infrastructure and cardiology specialists in geographically large areas and difficulty in accessing health services. In addition, the current COVID-19 pandemic poses new difficulties in the clinical evaluation of patients by delaying important in-person patient-doctor contacts, negatively impacting screening and monitoring activities.
\n\nA non-invasive assessment of the mechanical function of the heart, performed at point-of-care settings, can provide early information regarding congenital and acquired heart diseases in children. The lack of early diagnoses of these conditions represents a major public health problem, especially in underprivileged countries with high birth rates [5, 6, 7]. In particular, cardiac auscultation and the analysis of the phonocardiogram (PCG) can unveil fundamental clinical information regarding heart malfunctioning caused by congenital and acquired heart disease in pediatric populations. This is achieved by detecting abnormal sound waves, or heart murmurs, in the PCG signal. Murmurs are abnormal waves generated by turbulent blood flow in cardiac and vascular structures. They are closely associated with specific diseases such as septal defects, failure of ductus arteriosus closure in newborns, and defective cardiac valves.
\n\nSuccedent the 2016 Challenge, which focused on classifying normal vs. abnormal heart sounds from a single short recording from a single precordial location [8, 9], this year’s Challenge is devoted to detecting the presence or absence of murmurs from multiple heart sound recordings from multiple auscultation locations, as well as detecting the clinical outcomes.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":1361419899,"compressed_storage_size":1354550382,"resource_type":"Challenge","access_policy":"Open","topics":["challenge","competition","cardiac auscultation","congenital heart diseases"],"source_url":"https://physionet.org/content/challenge-2022/1.0.0/"},{"public_project_uuid":"bab045e8-54b5-4c20-82aa-d79ec743b468","slug":"inspire","version":"1.0","core_doi":"10.13026/dv8j-5412","version_doi":"10.13026/jyzb-ez61","is_latest_version":false,"short_description":"A public dataset that contains information related to surgery, anesthesia, laboratory results, medications, diagnosis, and outcomes from 50% of the patients who received surgery at Seoul National University Hospital between 2011 and 2020.","publish_date":"2023-10-06","title":"INSPIRE, a publicly available research dataset for perioperative medicine","abstract":"We present the INSPIRE dataset, a publicly available research dataset in perioperative medicine, which includes approximately 130,000 cases (50% of all surgical cases) who underwent anesthesia for surgery at an academic institution in South Korea between 2011 and 2020. This comprehensive dataset includes patient characteristics such as age, sex, American Society of Anesthesiologists physical status classification, diagnosis, surgical procedure code, department, and type of anesthesia. It also includes vital signs in the operating theatre, general wards, and intensive care units (ICUs), laboratory results from six months before admission to six months after discharge, and medication during hospitalization. Complications include total hospital and ICU length of stay and in-hospital death. We hope this dataset will inspire collaborative research and development in perioperative medicine and serve as a reproducible external validation dataset to improve surgical outcomes.
","license":{"name":"Korea Credentialed Health Data License-1-0-0"},"dua":{"name":"Korea Credentialed Health Data Agreement-1-0-0"},"main_storage_size":4689024633,"compressed_storage_size":579988667,"resource_type":"Database","access_policy":"Credentialed","topics":["perioperative medicine","multi-center","open dataset","surgery"],"source_url":"https://physionet.org/content/inspire/1.0/"},{"public_project_uuid":"3201cad5-361b-4d0d-8cb1-620eeb33f3e2","slug":"icehg-ds","version":"1.0.1","core_doi":"10.13026/vrvd-mf28","version_doi":"10.13026/zw34-n382","is_latest_version":true,"short_description":"The design and development of ICEHG DS was funded by the Slovenian Research Agency (ARRS) under the research project Metabolic and inborn factors of reproductive health, birth III.","publish_date":"2023-10-08","title":"Induced Cesarean EHG DataSet (ICEHG DS): An open dataset with electrohysterogram records of pregnancies ending in induced and cesarean section delivery","abstract":"The surface electrohysterogram (EHG) has emerged as a promising diagnostic tool for non-invasive automated preterm birth prediction. The existing automated preterm birth prediction methods rely on the use of uterine EHG records coming from spontaneous preterm and term deliveries, and are indifferent to term induced and cesarean deliveries. In order to enhance current publicly available pool of term EHG records, we developed a new EHG dataset containing 126 30-minute EHG records, recorded either early (23rd week), or later during pregnancy (31st week), of those pregnancies that were expected to end in spontaneous term delivery, but ended in induced or cesarean section delivery. The records were collected at the University Medical Center Ljubljana, Ljubljana, Slovenia. The dataset includes 38 and 43, early and later, induced; 11 and 8, early and later, cesarean; and 13 and 13, early and later, induced-cesarean records. This database enables better understanding of the underlying physiological mechanisms involved during pregnancies ending in induced and cesarean deliveries, and provides more robust and realistic assessment of the performance of automated preterm birth prediction.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":121177789,"compressed_storage_size":93277918,"resource_type":"Database","access_policy":"Open","topics":["cesarean-section delivery","neuroelectric","induced delivery","pregnancy","electrohysterogram"],"source_url":"https://physionet.org/content/icehg-ds/1.0.1/"},{"public_project_uuid":"5cb822ed-4c6d-413e-88f6-99e144fe33ec","slug":"carmen-i","version":"1.0","core_doi":"10.13026/bkwd-3j50","version_doi":"10.13026/bxrx-y344","is_latest_version":false,"short_description":"CARMEN-I is a Spanish corpus of 2,000 clinical records from Hospital Clínic, Barcelona. It covers COVID-19 patients and comorbidities, serving as a resource for training clinical NLP models and researchers in NLP applied to clinical documents.","publish_date":"2023-11-02","title":"CARMEN-I: A resource of anonymized electronic health records in Spanish and Catalan for training and testing NLP tools","abstract":"The CARMEN-I corpus comprises 2,000 clinical records, encompassing discharge letters, referrals, and radiology reports from Hospital Clínic of Barcelona between March 2020 and March 2022. These reports, primarily in Spanish with some Catalan sections, cover COVID-19 patients with diverse comorbidities like kidney failure, cardiovascular diseases, malignancies, and immunosuppression. The corpus underwent thorough anonymization, validation, and expert annotation, replacing sensitive data with synthetic equivalents. A subset of the corpus features annotations of medical concepts by specialists, encompassing symptoms, diseases, procedures, medications, species, and humans (including family members). CARMEN-I serves as a valuable resource for training and assessing clinical NLP techniques and language models, aiding tasks like de-identification, concept detection, linguistic modifier extraction, document classification, and more. It also facilitates training researchers in clinical NLP and is a collaborative effort involving Barcelona Supercomputing Center's NLP4BIA team, Hospital Clínic, and Universitat de Barcelona's CLiC group.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":23617213,"compressed_storage_size":16267038,"resource_type":"Database","access_policy":"Contributor Review","topics":["de-identification","anonymization","clinical ner"],"source_url":"https://physionet.org/content/carmen-i/1.0/"},{"public_project_uuid":"bcba7c48-82ad-4e5c-96d6-3e5309feda0d","slug":"transformer-deid","version":"1.0.0","core_doi":"10.13026/gyp9-4164","version_doi":"10.13026/7dj5-7x85","is_latest_version":true,"short_description":"Fine tune transformer-based neural networks to deidentify clinical text data.","publish_date":"2023-11-02","title":"Transformer-DeID: Deidentification of free-text clinical notes with transformers","abstract":"Sharing patient data such as clinical notes among clinicians and researchers is fundamental to advancements in clinical practice and biomedical research. Prior to distribution, it is often necessary and legally mandated to remove identifiers such as names, contact details, identification numbers, and dates to safeguard patient privacy. This process is known as deidentification. Here we provide a neural network model for the removal of patient identifiers from clinical text based upon Bidirectional Encoder Representations from Transformers (BERT) and two variations of BERT, namely Robustly Optimized BERT Approach (RoBERTa) and Distilled BERT (DistilBERT). The models demonstrated excellent performance on a publicly available benchmark and are made freely available to maximize reuse.
","license":{"name":"MIT License"},"dua":null,"main_storage_size":1188235127,"compressed_storage_size":1083415115,"resource_type":"Software","access_policy":"Open","topics":["neural networks","transformers","deidentification"],"source_url":"https://physionet.org/content/transformer-deid/1.0.0/"},{"public_project_uuid":"30d43f05-5149-4b23-bbef-9e047a9446d3","slug":"corpus-fungal-infections","version":"1.0.1","core_doi":"10.13026/g3t3-2c54","version_doi":"10.13026/ssk0-1x59","is_latest_version":false,"short_description":"A corpus of cytology and histopathology reports annotated for terminology relevant to fungal infections. Ideal for validation of named entity recognition and relation extraction methods.","publish_date":"2023-11-02","title":"CHIFIR: Cytology and Histopathology Invasive Fungal Infection Reports","abstract":"Surveillance of invasive fungal infection (IFI) in clinical settings is a laborious process requiring a detailed review of patient medical history. One of the key sources of clinical information is cytology and histopathology reports: pathologist-produced free-text reports outlining the macroscopic and microscopic structure of a specimen. The data was generated to facilitate the development of an automated tool for the detection of IFI. The Cytology and Histopathology Invasive Fungal Infection Reports (CHIFIR) corpus contains 283 de-identified reports annotated by infectious diseases physicians for terminology relevant to the IFI diagnosis. These include IFI-specific concepts and certainty cues, as well as relations between them. We release the annotation schema and the original reports along with the corresponding annotation files. We anticipate the CHIFIR corpus to be useful in the development and validation of named entity recognition and relation extraction methods with a focus on clinical data. Such methods can be instrumental in processing other types of clinical documentation (radiology reports, clinical notes, nursing notes) with various downstream tasks in mind.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":504941,"compressed_storage_size":401524,"resource_type":"Database","access_policy":"Credentialed","topics":["clinical documentation","information extraction","nlp","invasive fungal infections"],"source_url":"https://physionet.org/content/corpus-fungal-infections/1.0.1/"},{"public_project_uuid":"5d9ff82c-7222-4bb6-919a-24b5dbab008a","slug":"inspire","version":"1.1","core_doi":"10.13026/dv8j-5412","version_doi":"10.13026/gjyw-ea42","is_latest_version":false,"short_description":"A public dataset that contains information related to surgery, anesthesia, laboratory results, medications, diagnosis, and outcomes from 50% of the patients who received surgery at Seoul National University Hospital between 2011 and 2020.","publish_date":"2023-11-03","title":"INSPIRE, a publicly available research dataset for perioperative medicine","abstract":"We present the INSPIRE dataset, a publicly available research dataset in perioperative medicine, which includes approximately 130,000 cases (50% of all surgical cases) who underwent anesthesia for surgery at an academic institution in South Korea between 2011 and 2020. This comprehensive dataset includes patient characteristics such as age, sex, American Society of Anesthesiologists physical status classification, diagnosis, surgical procedure code, department, and type of anesthesia. It also includes vital signs in the operating theatre, general wards, and intensive care units (ICUs), laboratory results from six months before admission to six months after discharge, and medication during hospitalization. Complications include total hospital and ICU length of stay and in-hospital death. We hope this dataset will inspire collaborative research and development in perioperative medicine and serve as a reproducible external validation dataset to improve surgical outcomes.
","license":{"name":"Korea Credentialed Health Data License-1-0-0"},"dua":{"name":"Korea Credentialed Health Data Agreement-1-0-0"},"main_storage_size":4767346995,"compressed_storage_size":557654018,"resource_type":"Database","access_policy":"Credentialed","topics":["perioperative medicine","multi-center","open dataset","surgery"],"source_url":"https://physionet.org/content/inspire/1.1/"},{"public_project_uuid":"a16d93c2-fd88-4f78-b2ea-27a95ecf4c07","slug":"stigmatizing-language","version":"1.0.0","core_doi":"10.13026/n938-2385","version_doi":"10.13026/53xq-jz70","is_latest_version":true,"short_description":"A suite of classifiers for detecting three types of stigmatizing language in electronic medical records. Trained on MIMIC-IV discharge notes.","publish_date":"2023-11-06","title":"Characterization of Stigmatizing Language in Medical Records","abstract":"Widespread disparities in clinical outcomes exist between different demographic groups in the United States. A new line of work in medical sociology has demonstrated physicians often use stigmatizing language in electronic medical records within certain groups, such as black patients, which may exacerbate disparities. The first step to addressing the presence of stigmatizing language in medical records is identifying it and characterizing its impact. Towards this end, we release a suite of neural and non-neural classifiers trained in a supervised manner to recognize three types of stigmatizing language found in discharge notes from the MIMIC-IV dataset. We also release the set of 5,043 annotations from 4,710 notes (4,259 patients) used to train and evaluate our models. These resources provide a foundation for NLP researchers to contribute timely insights to a problem domain brought to the forefront by recent legislation regarding clinical documentation transparency.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":2615896445,"compressed_storage_size":2420921582,"resource_type":"Model","access_policy":"Credentialed","topics":["clinical natural language processing","domain transfer","bias","stigmatizing language","mimic","large language models"],"source_url":"https://physionet.org/content/stigmatizing-language/1.0.0/"},{"public_project_uuid":"986da9ee-75f7-4de2-8b62-68d8ff81bf76","slug":"blood-gas-oximetry","version":"1.0","core_doi":"10.13026/vfba-0n65","version_doi":"10.13026/phvt-3277","is_latest_version":true,"short_description":"An open-source pulse oximetry and arterial blood gas dataset, derived from MIMIC-III, MIMIC-IV, and eICU-CRD","publish_date":"2023-11-09","title":"BOLD, a blood-gas and oximetry linked dataset","abstract":"Pulse oximeters measure peripheral arterial oxygen saturation (SpO2) noninvasively, while the gold standard involves arterial blood gas measurement (SaO2). There are known racial and ethnic disparities in their performance. BOLD is a new comprehensive dataset that aims to underscore the importance of addressing biases in pulse oximetry accuracy, particularly affecting people with darker skin tones. The dataset was created by harmonizing three Electronic Health Record databases (MIMIC-III, MIMIC-IV, eICU-CRD) comprising Intensive Care Unit stays. Paired SpO2 and SaO2 measurements were time-aligned and combined with various other clinical parameters to provide a detailed picture of each patient. It includes 49,099 such paired measurements within a 5-minute window, with oxygen saturation levels between 70-100%. Significantly, about 25% of the data represents minority racial and ethnic groups, a proportion seldom achieved in previous studies. The code scripts have been made publicly available to facilitate replication.We hope that BOLD is poised to be a valuable resource for health equity studies, that can be used to develop pulse oximetry debiasing algorithms.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":32687913,"compressed_storage_size":9305971,"resource_type":"Database","access_policy":"Credentialed","topics":["health equity","electronic health records","pulse oximetry","intensive care unit"],"source_url":"https://physionet.org/content/blood-gas-oximetry/1.0/"},{"public_project_uuid":"cafcec43-0e89-44d3-86fe-25359caea609","slug":"respiratory-dataset","version":"1.0.0","core_doi":"10.13026/psrj-rw40","version_doi":"10.13026/d767-e709","is_latest_version":true,"short_description":"Outlined is a pressure, flow, volume, dynamic circumference, and EIT assessed aeration dataset from resting breathing with REO at increasing CPAP PEEP settings. Vapers, asthmatics, smokers, and otherwise healthy people were included in the trial.","publish_date":"2023-11-10","title":"Respiratory dataset from PEEP study with expiratory occlusion","abstract":"A trial was conducted to collect gauge pressure and flow from a venturi-based flow-meter, dynamic abdominal and thoracic circumference from rotary-encoder based tape measures, and aeration data from electrical impedance tomography (EIT). Data was collected from 80 adults, breathing with continuous positive airway pressure (CPAP) ventilation, under ethical consent from the University of Canterbury Human Research Ethics Committee (HREC 2023/04/LR-PS). In each subject’s recording the positive end expiratory pressure (PEEP) was increased from 4 to 12 cmH2O in 0.5 cmH2O increments, with data recorded for 30 seconds at each level. Time was built in for researchers to change PEEP settings, in which data continued to be collected. The recording also begins and ends with a 60 second period of breathing without CPAP. During the trial subjects breathed through a full-face mask and filter collected to the data collection device and, when PEEP was applied, CPAP circuitry. A camera-shutter based device was used to rapidly occlude the expiratory pathway to enable identification of passive lung mechanics. Subjects were instructed to breathe normally throughout the trial. Both raw and processed data is included in this publication to maximise its utility. Subject demographic data was self-reported using a questionnaire completed prior to each trial and is collated in a spreadsheet as part of this dataset. The demographic data collected was as follows: sex; height; weight; age; any history of asthma; smoking, or vaping; and resting chest width and depth. Ultimately, this dataset was collected to enable the development and validation of model-based respiratory function assessment methods. These methods aim to increase the capacity of automated testing, removing clinical burden of respiratory monitoring and improving patient-specific care.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":19619404137,"compressed_storage_size":15213230928,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/respiratory-dataset/1.0.0/"},{"public_project_uuid":"a48b3155-963a-4707-882a-82019a7d5ecc","slug":"bionlp-workshop-2023-task-1a","version":"2.0.0","core_doi":"10.13026/s8wk-ja78","version_doi":"10.13026/mg5c-x807","is_latest_version":true,"short_description":"This is the data storage for BioNLP Workshop Shared Task 1A: Problem List Summarization.","publish_date":"2023-11-12","title":"BioNLP Workshop 2023 Shared Task 1A: Problem List Summarization","abstract":"Automatically summarizing patients’ main problems from the daily care notes in the electronic health record can help mitigate information and cognitive overload for clinicians and provide augmented intelligence via computerized diagnostic decision support at the bedside. The task of Problem List Summarization aims to generate a list of diagnoses and problems in a patient’s daily care plan using input from the provider’s progress notes during hospitalization.
\n\nThis task aims to promote NLP model development for downstream applications in diagnostic decision support systems that could improve efficiency and reduce diagnostic errors in hospital care. The task contains 768 hospital daily progress notes and 2783 diagnoses in the training set, and a new set of 237 daily progress notes are recently annotated as the test set. The annotation methods and annotation quality have previously been reported.
\n\nThe dataset supports a more complex summarization task to generate a list of relevant diagnoses/problems given the information in the Subjective, Objective, and Assessment sections of the note. Only diagnoses/problems that are available in the progress note were labeled for the task.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":2868767,"compressed_storage_size":882753,"resource_type":"Challenge","access_policy":"Credentialed","topics":["bionlp","clinical natural language processing","electronic health record summarization"],"source_url":"https://physionet.org/content/bionlp-workshop-2023-task-1a/2.0.0/"},{"public_project_uuid":"d4c21c0b-1a2a-48ec-9f34-bfca912f5e24","slug":"scientisst-move-biosignals","version":"1.0.0","core_doi":"10.13026/0ppk-ha30","version_doi":"10.13026/sg89-qq52","is_latest_version":false,"short_description":"Multimodal (ECG, EMG, EDA, PPG, TEMP, ACC) biosignal dataset of everyday activities. Created with 3 wearable devices based on ScientISST Sense and Empatica E4.","publish_date":"2023-11-13","title":"ScientISST MOVE: Annotated Wearable Multimodal Biosignals recorded during Everyday Life Activities in Naturalistic Environments","abstract":"Physiological processes are affected and responsive to daily stimuli, and their daily monitoring can be key to understanding well-being and detecting diseases. Since most physiological data is acquired at rest, or in controlled scenarios, many decision algorithms based on physiological data may not perform as well during free-living conditions. ScientISST MOVE is a 3-wearable biosignal dataset recorded during natural everyday activities, including lifting a chair, greeting, gesticulating, walking and running. The wearable devices: a chestband, an armband, and the Empatica E4 wristband, recorded a double view of electrocardiography (ECG), electrodermal activity (EDA), and photoplethysmography (PPG). Additionally, they also recorded bicep electromyography (EMG), wrist temperature (TEMP) and chest and wrist actigraphy (ACC). 17 healthy volunteers participated in the data collection, resulting in a total of 10.5 sensor-simultaneous useful hours. ScientISST MOVE enables the study of the full body state in free-living environments, and the effect of daily activities on physiology. Moreover, the double view of physiological data herein provided could be used in signal quality comparison studies.
","license":{"name":"Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":468595499,"compressed_storage_size":206592651,"resource_type":"Database","access_policy":"Open","topics":["run","uncontrolled environments","jump","multimodal","greet","lift","walk","wearable","gesticulate"],"source_url":"https://physionet.org/content/scientisst-move-biosignals/1.0.0/"},{"public_project_uuid":"7f323d0f-2115-4c9f-9d08-bcb71e9ca32b","slug":"simulated-obstructive-disease","version":"1.0.0","core_doi":"10.13026/v59r-fy88","version_doi":"10.13026/xczc-3662","is_latest_version":true,"short_description":"Outlined is a pressure, flow, and volume dataset using a using a modular device to simulate the effects of obstructive pulmonary disease in healthy people. 20 healthy subjects were included in this dataset.","publish_date":"2023-11-13","title":"Simulated Obstructive Disease Respiratory Pressure and Flow","abstract":"The dataset presented is from 20 healthy adults at the University of Canterbury, where the tests were approved by its Human Research Ethics Committee (HREC) ethics. Data was collected using a modular device simulating the effects of Chronic Obstructive Pulmonary Disease (COPD) in healthy people, and in particular, the gas trapping due to airway collapse in expiration. Pressure and flow data were recorded in this trial using a y-split venturi with the device attached to the exhalatory outlet by the mouth. Data was collected using continuous positive airway pressure (CPAP) ventilation at positive end-expiratory pressure (PEEP) settings of 4 and 8 cmH20, and without the CPAP at zero end-expiratory pressure (ZEEP). The recording time for each test was 45 seconds with time in between each test to allow subjects to recover. Each subject filled out a de-identified demographic questionnaire to gather background information, including age, height, sex, and history of smoking, vaping, or asthma. The data collected is intended to be used as a preliminary validation method of respiratory computational models, where the device use avoids the need to test unhealthy people or use invasive methods.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":160445741,"compressed_storage_size":35864242,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/simulated-obstructive-disease/1.0.0/"},{"public_project_uuid":"614192c7-6896-4bef-bc05-697bf2250c59","slug":"annotation-dataset-sdoh","version":"1.0.0","core_doi":"10.13026/p8zc-cs44","version_doi":"10.13026/6149-mb25","is_latest_version":false,"short_description":"Annotation dataset of social determinants of health from MIMC-III Clinical Care Database notes.","publish_date":"2023-11-24","title":"Annotation dataset of social determinants of health from MIMIC-III Clinical Care Database","abstract":"Social determinants of health (SDoH) have an important impact on patient outcomes but are incompletely collected from the electronic health records (EHR). This study researched the ability of large language models to extract SDoH from free text in EHRs, where they are most commonly documented, and explored the role of synthetic clinical text for improving the extraction of these scarcely documented, yet extremely valuable, clinical data. We developed annotation guidelines for sentence-level annotation of SDoH that are not reliably available as structured data in the EHR: employment, housing, transportation, parental status, relationship, and social support. Sentences were labeled for both the presence of an SDoH mention and the presence of an adverse SDoH mention. After finalizing the annotation guidelines, two annotators manually annotated a separate corpus, which cannot be released due to PHI. A total of 300/800 (37.5%) of these notes underwent dual annotation. Before adjudication, dually-annotated notes had a Krippendorf’s alpha agreement of 0.86 and Cohen’s Kappa of 0.86 for any SDoH mention categories. For adverse SDoH mentions, notes had a Krippendorf’s alpha agreement of 0.76 and Cohen’s Kappa of 0.76. As an external validation, 200 notes from MIMIC-III written by physicians, social workers, and nurses were manually annotated by a single annotator. Here, we release this manually annotated corpus of 200 MIMC-III notes.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1760080,"compressed_storage_size":530477,"resource_type":"Database","access_policy":"Credentialed","topics":["natural language processing","social determinants of health"],"source_url":"https://physionet.org/content/annotation-dataset-sdoh/1.0.0/"},{"public_project_uuid":"f63c0d4a-609e-4099-8894-14921d16fc6e","slug":"cad-chest","version":"1.0","core_doi":"10.13026/kyfs-5608","version_doi":"10.13026/44pd-vz36","is_latest_version":true,"short_description":"The CAD-Chest dataset provides comprehensive annotations of disease, including disease severity, uncertainty, and location based on the MIMIC-CXR radiologist reports.","publish_date":"2023-12-08","title":"CAD-Chest: Comprehensive Annotation of Diseases based on MIMIC-CXR Radiology Report","abstract":"Several extant chest X-ray (CXR) datasets predominantly comprise binary disease labels and exhibit a deficiency in providing comprehensive disease-related information. Crucial facets of disease management, including disease severity, diagnostic uncertainty, and precise localization, are often absent in these datasets, yet they hold substantial clinical significance. In this work, we present a comprehensive annotation of disease (CAD) on CXR images, which is named CAD-Chest dataset. We have leveraged radiology reports authored by medical professionals to meticulously devise label extraction protocols. These protocols facilitate the extraction of essential disease-related attributes, encompassing disease name, severity grading, and additional pertinent details. This dataset is poised to empower researchers and practitioners by offering a holistic perspective on diseases, transcending the mere presence or absence of binary classification.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":358252846,"compressed_storage_size":15601156,"resource_type":"Database","access_policy":"Credentialed","topics":["chesr x-ray","disease label"],"source_url":"https://physionet.org/content/cad-chest/1.0/"},{"public_project_uuid":"8a442bfb-8747-4bf8-9c36-8efd512bedfa","slug":"medication-labels-mimic-note","version":"1.0.0","core_doi":"10.13026/6d3r-w470","version_doi":"10.13026/ps1s-ab29","is_latest_version":true,"short_description":"Medication extraction NLP labels for 600 discharge summaries in MIMIC-IV-Note dataset.","publish_date":"2023-12-12","title":"Medication Extraction Labels for MIMIC-IV-Note Clinical Database","abstract":"This dataset release provides medication extraction labels for a subset of 600 discharge summaries from the 2023 MIMIC-IV-Note dataset. These labels are consistent with the schema from the 2009 i2b2 Workshop on NLP Challenges dataset. We utilized a Large Language Model (LLM) pipeline to generate these labels, achieving performance on par with the average human annotation specialist.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1925378,"compressed_storage_size":909982,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/medication-labels-mimic-note/1.0.0/"},{"public_project_uuid":"8ed9468a-4578-48ab-b26d-a056d2713afa","slug":"i-care","version":"2.1","core_doi":"10.13026/sd64-2856","version_doi":"10.13026/m33r-bj81","is_latest_version":true,"short_description":"The clinical and EEG data for this dataset originates from seven academic hospitals in the U.S. and Europe led by investigators part of the International Cardiac Arrest REsearch consortium (I-CARE).","publish_date":"2023-12-15","title":"I-CARE: International Cardiac Arrest REsearch consortium Database","abstract":"The International Cardiac Arrest REsearch consortium (I-CARE) Database includes baseline clinical information and continuous electroencephalogram (EEG) and electrocardiogram (ECG) recordings from comatose patients following cardiac arrest. The patients were admitted to an intensive care unit (ICU) in one of seven academic hospitals in the U.S. and Europe and monitored for several hours to several days. The long-term neurological function of the patients was determined using the Cerebral Performance Category scale.
","license":{"name":"Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":1693972610522,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/i-care/2.1/"},{"public_project_uuid":"3d60e8b6-11b8-47c8-b06a-db139cccc177","slug":"kiel-cardio","version":"1.0.0","core_doi":"10.13026/16zn-4m55","version_doi":"10.13026/q6r2-zz68","is_latest_version":true,"short_description":"The Kiel Cardio Database (KCD) contains one-minute 8-lead magnetocardiographic (MCG) measurements from seven subjects. Each subject underwent 25 consecutive measurements using a sensor array comprising four QuSpin QZFMs.","publish_date":"2023-12-15","title":"Kiel Cardio Database","abstract":"The Kiel Cardio Database, which includes magnetocardiography (MCG) measurements of seven healthy male volunteers, is presented in this publication. QuSpin QZFM Gen 1 sensors were used in a magnetically shielded chamber at the University of Kiel. Each subject underwent 25 one-minute measurements, resulting in magnetic field recordings at 200 different positions for each subject. The dataset is intended to facilitate MCG research by making this expensive method accessible to researchers without the necessary equipment. Future versions of the database will expand its utility for cardiac research and arrhythmogenic tissue localization by incorporating electrocardiography (ECG), magneto-tomographic imaging (MRI), and annotation data.
","license":{"name":"Creative Commons Attribution-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":7301594032,"compressed_storage_size":7007649503,"resource_type":"Database","access_policy":"Open","topics":["opically pumped magnetometer","opm","mcg","magnetocardiography"],"source_url":"https://physionet.org/content/kiel-cardio/1.0.0/"},{"public_project_uuid":"dc2866d8-e33b-4779-9c11-26ef96c66615","slug":"brain-wearable-monitoring","version":"1.0.0","core_doi":"10.13026/5yxs-9t42","version_doi":"10.13026/jr6f-wc85","is_latest_version":true,"short_description":"This dataset explores safe actuation (music, perfume, coffee) for enhancing cognitive states. It includes subjects’ responses, reaction times, and physiological data (EDA, HR, BVP, PPG, temperature, accelerometer, EEG) during n-back memory tasks.","publish_date":"2023-12-18","title":"Regulation of Brain Cognitive States through Auditory, Gustatory, and Olfactory Stimulation with Wearable Monitoring","abstract":"Inspired by advances in wearable technologies, we design and perform human-subject experiments to investigate the impacts of applying safe actuation (i.e., auditory, gustatory, and olfactory) for the purpose of regulating the cognitive arousal state and enhancing the performance state. In two separate experiments, participants are instructed to engage in a working memory task known as the n-back task. To modulate their brain state, we introduce various interventions, including listening to different genres of music in the first experiment, and consuming coffee and smelling perfume in the second experiment. Employing only wearable devices for human monitoring and using safe actuation intervention are the key components of the performed experiments. This dataset contains subjects' correct/incorrect responses along with their reaction times. In these experiments, we employed two Empatica wristbands and one muse headband to collect their physiological data. The data from Empatica contains participants' electrodermal activity (EDA), heart rate (HR), blood volume pulses (BVP), skin surface temperature, Photoplethysmography (PPG), and 3-axis accelerometer data. Using a muse headband, we also collected Electroencephalography (EEG) signals from four sensors. By creating a dataset that addresses the current lack of publicly available data for utilizing wearable devices and safe everyday stimuli to manage internal brain states, our research enables further investigations into machine learning and system identification. It eventually paves the way for smarter work environments in the future. The ultimate goal is to develop practical and automated personalized closed-loop systems that can effectively regulate internal brain states and enhance the overall quality of life.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":2938222946,"compressed_storage_size":585221370,"resource_type":"Database","access_policy":"Open","topics":["signal processing","neurophysiology","coffee","music","empatica","n-back experiments","perfume","physiological signals","wearable devices","control systems","eeg","memory task","heart rate","muse","closed-loop","brain cognitive states","skin conductance"],"source_url":"https://physionet.org/content/brain-wearable-monitoring/1.0.0/"},{"public_project_uuid":"3d1cc17a-0486-4da3-96f9-354be66e60ad","slug":"snomed-ct-entity-challenge","version":"1.0.0","core_doi":"10.13026/3f69-tk39","version_doi":"10.13026/s48e-sp45","is_latest_version":false,"short_description":"272 discharge notes from the MIMIC-IV-Note dataset annotated with SNOMED CT concepts.","publish_date":"2023-12-19","title":"SNOMED CT Entity Linking Challenge","abstract":"This challenge, sponsored by SNOMED International, seeks to advance the development of Entity Linking models that operate on unstructured clinical texts. Participants in the challenge will train entity linking models using a subset of MIMIC-IV-Note discharge summaries that have been annotated with SNOMED CT concepts by a team of medical professionals. The full dataset (which is comprised of a training set and a test set) consists of approximately 75,000 annotations across nearly 300 discharge summaries. The competition is being hosted by DrivenData, whose platform will manage registration, code submission, and evaluation.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":3720252,"compressed_storage_size":1125259,"resource_type":"Challenge","access_policy":"Credentialed","topics":["snomed","entity linking","clinical annotation"],"source_url":"https://physionet.org/content/snomed-ct-entity-challenge/1.0.0/"},{"public_project_uuid":"4dc44982-93b2-4fe8-bb52-b14841e2c2d7","slug":"inspire","version":"1.2","core_doi":"10.13026/dv8j-5412","version_doi":"10.13026/4evs-wq50","is_latest_version":false,"short_description":"A public dataset that contains information related to surgery, anesthesia, laboratory results, medications, diagnosis, and outcomes from 50% of the patients who received surgery at Seoul National University Hospital between 2011 and 2020.","publish_date":"2023-12-28","title":"INSPIRE, a publicly available research dataset for perioperative medicine","abstract":"We present the INSPIRE dataset, a publicly available research dataset in perioperative medicine, which includes approximately 130,000 cases (50% of all surgical cases) who underwent anesthesia for surgery at an academic institution in South Korea between 2011 and 2020. This comprehensive dataset includes patient characteristics such as age, sex, American Society of Anesthesiologists physical status classification, diagnosis, surgical procedure code, department, and type of anesthesia. It also includes vital signs in the operating theatre, general wards, and intensive care units (ICUs), laboratory results from six months before admission to six months after discharge, and medication during hospitalization. Complications include total hospital and ICU length of stay and in-hospital death. We hope this dataset will inspire collaborative research and development in perioperative medicine and serve as a reproducible external validation dataset to improve surgical outcomes.
","license":{"name":"Korea Credentialed Health Data License-1-0-0"},"dua":{"name":"Korea Credentialed Health Data Agreement-1-0-0"},"main_storage_size":4657153574,"compressed_storage_size":535976548,"resource_type":"Database","access_policy":"Credentialed","topics":["multi-center","surgery","perioperative medicine","open dataset"],"source_url":"https://physionet.org/content/inspire/1.2/"},{"public_project_uuid":"d6850b20-4cdb-4fee-869e-aff90b55768a","slug":"hd-semg","version":"2.0.0","core_doi":"10.13026/t9wv-d929","version_doi":"10.13026/hxan-pe94","is_latest_version":true,"short_description":"We provide an open access dataset of High Density Surface Electromyogram (HD-sEMG). Our dataset can be used for both hand gesture classification-based neuroprosthetic control and EMG-force regression based proportional neuroprosthetic control.","publish_date":"2023-12-28","title":"Open Access Dataset and Toolbox of High-Density Surface Electromyogram Recordings","abstract":"We provide an open access dataset of High densitY Surface Electromyogram (HD-sEMG) Recordings (named "Hyser"). We acquired data from 20 subjects with each subject participating in our experiment twice on separate days following the same experiment paradigm. Our Hyser dataset contains five sub-datasets: (1) pattern recognition (PR) dataset acquired during 34 hand gestures, (2) maximal voluntary muscle contraction (MVC) dataset while subjects contracted each individual finger, (3) one-degree of freedom (DoF) dataset acquired during force-varying contraction of each individual finger, (4) N-DoF dataset acquired during prescribed contractions of combinations of multiple fingers, and (5) random task dataset acquired during random contraction of combinations of fingers without any prescribed force trajectory. Sub-dataset 1 can be used for gesture recognition studies. Sub-datasets 2-5 also recorded individual finger forces, thus can be used for studies on proportional control of neuroprostheses.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":153288112758,"compressed_storage_size":145134191274,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/hd-semg/2.0.0/"},{"public_project_uuid":"83face4b-9889-4f98-b81c-f9cc314172b5","slug":"patient-level-data-covid-ms","version":"1.0.1","core_doi":"10.13026/3cmh-xa68","version_doi":"10.13026/77ta-1866","is_latest_version":true,"short_description":"This dataset is part of the Global Data Sharing Initiative. The data was acquired by people with MS and clinicians using a fast data entry tool. The dataset includes demographics, comorbidities and hospital stay and COVID-19 symptoms of PwMS.","publish_date":"2024-01-02","title":"Patient-level dataset to study the effect of COVID-19 in people with Multiple Sclerosis","abstract":"Multiple Sclerosis (MS) is an inflammatory autoimmune disease of the central nervous system, causing increased vulnerability to infections and disability among young adults. Ever since the coronavirus disease 2019 (COVID-19) outbreak, caused by severe acute respiratory syndrome coronavirus 2 infections, there have been concerns among people with MS (PwMS) about the potential interactions between various disease-modifying therapies and COVID-19. The COVID-19 in MS Global Data Sharing Initiative (GDSI) was initiated in 2020 to address these concerns. This paper focuses on the anonymisation and open-sourcing of a GDSI sub-dataset, comprising data entered by people with MS and clinicians using a fast data entry tool. The dataset includes demographics, comorbidities, hospital stay, and COVID-19 symptoms of PwMS. The dataset can be used to perform different statistical analyses to improve our understanding of COVID-19 in MS. Furthermore, this dataset can also be used within the context of educational activities to educate different stakeholders on the complex data science topics that were used within the GDSI.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":243074,"compressed_storage_size":23939,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/patient-level-data-covid-ms/1.0.1/"},{"public_project_uuid":"5e825ffe-0504-4de4-a072-95f2037e9fbe","slug":"neurocritical-pediatric","version":"1.0.0","core_doi":"10.13026/n8k7-bh29","version_doi":"10.13026/5kvn-pp29","is_latest_version":true,"short_description":"The database contains waveform recordings, including arterial blood pressure, intracranial pressure, and cerebral blood flow velocity, from pediatric patients in neurocritical care.","publish_date":"2024-01-09","title":"Neurocritical care waveform recordings in pediatric patients","abstract":"This database contains de-identified, time-locked waveform recordings from pediatric patients in neurocritical care. Typical waveforms collected at the bedside include arterial blood pressure, intracranial pressure, and cerebral blood flow velocity. Additionally, information include the hematocrit and the vertical location (i.e. height) of the blood pressure transducer and intracranial pressure transducer to account for potential hydrostatic pressure differences between the two pressure measurements. In total, the initial release contains data from 12 patients with a variety of pathologies necessitating invasive intracranial pressure monitoring. The total recording duration is about 10 hours across all patients and studies.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":47341260,"compressed_storage_size":25405136,"resource_type":"Database","access_policy":"Credentialed","topics":["arterial blood pressure","noninvasive icp","neurocritical care","neurotrauma","cerebral blood flow velocity","pediatric patients","intracranial pressure"],"source_url":"https://physionet.org/content/neurocritical-pediatric/1.0.0/"},{"public_project_uuid":"c6c56e7a-d9e1-46a0-9790-9b32845c2e8e","slug":"asclepius-r","version":"1.0.0","core_doi":"10.13026/1wf7-pm14","version_doi":"10.13026/gwjr-7j25","is_latest_version":false,"short_description":"Asclepius: Publicly Available Clinical Large Language Models with Synthetic Clinical Notes Asclepius-R: A instruction-finetuned large language model with MIMIC-III clinical notes","publish_date":"2024-01-09","title":"Asclepius-R : Clinical Large Language Model Built On MIMIC-III Discharge Summaries","abstract":"The development of large language models tailored for handling patients’ clinical notes is often hindered by the limited accessibility and usability of these notes due to strict privacy regulations. To address these challenges, we first create synthetic large-scale clinical notes using publicly available case reports extracted from biomedical literature. We then use these synthetic notes to train our specialized clinical large language model, Asclepius. While Asclepius is trained on synthetic data, we assess its potential performance in real-world applications by evaluating it using real clinical notes. We benchmark Asclepius against several other large language models, including GPT-3.5-turbo and other open-source alternatives. To further validate our approach using synthetic notes, we also compare Asclepius with its variants trained on real clinical notes. Our findings convincingly demonstrate that synthetic clinical notes can serve as viable substitutes for real ones when constructing high-performing clinical language models. This conclusion is supported by detailed evaluations conducted by both GPT-4 and medical professionals. All resources—including weights, codes, and data—used in the development of Asclepius are made publicly accessible for future research. Specifically, this repository contains Asclepius-R, a variant of Asclepius that was trained on MIMIC-III discharge summaries. All other resource are also publicly accessible.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":79058055436,"compressed_storage_size":73163600170,"resource_type":"Model","access_policy":"Credentialed","topics":["clinical notes","large language model","synthetic clinical notes","open-source","clinical llm","asclepius","synthetic notes","llm"],"source_url":"https://physionet.org/content/asclepius-r/1.0.0/"},{"public_project_uuid":"2bce2524-2bd7-4f59-8645-86d79d54fcf1","slug":"semg","version":"1.0.1","core_doi":"10.13026/9q7v-6016","version_doi":"10.13026/grdj-qx06","is_latest_version":true,"short_description":"The dataset is composed of long-lasting surface electromyographic (sEMG) signals recorded from ten muscles during ground walking of 31 young able-bodied subjects in Movement Analysis Lab, Università Politecnica delle Marche, Ancona, Italy.","publish_date":"2024-01-09","title":"Surface electromyographic signals collected during long-lasting ground walking of young able-bodied subjects","abstract":"The present dataset is composed of long-lasting (around 5 minutes) surface electromyographic (sEMG) signals recorded from 2011 and 2018 during ground walking of 31 young (20 years < age < 30 years) able-bodied subjects in the Movement Analysis Lab, Università Politecnica delle Marche, Ancona, Italy. Underweight, overweight, and obese people (body mass index, BMI < 18.5 Kg/m2 and BMI > 25 Kg/m2) and subjects affected by any pathological condition, joint pain, or undergone orthopedic surgery are not included in the present dataset. sEMG signals are acquired from the following ten different leg muscles (five per leg): gastrocnemius lateralis (GL), tibialis anterior (TA), rectus femoris (RF), hamstrings (Ham), and vastus lateralis (VL). Synchronized footswitch and electrogoniometric data are provided in order to allow users to achieve a spatial/temporal characterization of the sEMG signals. Data have been acquired in subjects walking barefoot on level ground for around 5 min at their natural speed and pace, following an eight-shaped path which includes rectilinear segments and curves. The considerable length of the signals makes this dataset very suitable for those studies where the numerosity of the data is essential, such as machine/deep learning approaches, studies for analyzing and quantifying the variability of muscle recruitment during physiological walking, and creation of reference dataset in the characterization of pathological conditions.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":447782560,"compressed_storage_size":292006031,"resource_type":"Database","access_policy":"Open","topics":["biomedical signals","muscle recruitment","surface emg signal","walking","gait analysis"],"source_url":"https://physionet.org/content/semg/1.0.1/"},{"public_project_uuid":"7db43f25-45c6-411c-8a4b-c8ee9e70ce09","slug":"chexmask-cxr-segmentation-data","version":"0.3","core_doi":"10.13026/6eky-y831","version_doi":"10.13026/nv4g-fr21","is_latest_version":false,"short_description":"CheXmask Database is a 676,803 uniformly annotated chest radiographs with segmentation masks. Images were segmented using HybridGNet, with automatic quality control indicated by RCA scores.","publish_date":"2024-01-09","title":"CheXmask Database: a large-scale dataset of anatomical segmentation masks for chest x-ray images","abstract":"The CheXmask Database presents a comprehensive, uniformly annotated collection of chest radiographs, constructed from six public databases: CANDID-PTX, ChestX-ray8, Chexpert, MIMIC-CXR-JPG, Padchest and VinDr-CXR. The database aggregates 676,803 anatomical segmentation masks derived from images which have been processed using the HybridGNet model to ensure consistent, high-quality segmentation. To confirm the quality of the segmentations, we include in this database individual Reverse Classification Accuracy (RCA) scores for each of the segmentation masks. This dataset is intended to catalyze further innovation and refinement in the field of semantic chest X-ray analysis, offering a significant resource for researchers in the medical imaging domain.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":40457176541,"compressed_storage_size":16942093591,"resource_type":"Database","access_policy":"Open","topics":["chest x-ray segmentation","medical image segmentation","automatic quality assesment"],"source_url":"https://physionet.org/content/chexmask-cxr-segmentation-data/0.3/"},{"public_project_uuid":"ee7c964a-09af-48e3-8f1f-f12af2106667","slug":"ehr-ds-qa","version":"1.0.0","core_doi":"10.13026/9vd3-1y84","version_doi":"10.13026/25fx-f706","is_latest_version":true,"short_description":"Dataset consisting of question and answer pairs synthetically generated from medical discharge summaries, designed to facilitate the training and development of large language models specifically tailored for healthcare applications","publish_date":"2024-01-11","title":"EHR-DS-QA: A Synthetic QA Dataset Derived from Medical Discharge Summaries for Enhanced Medical Information Retrieval Systems","abstract":"This dataset was designed and created to enable advancements in healthcare-focused large language models, particularly in the context of retrieval-augmented clinical question-answering capabilities. Developed using a self-constructed pipeline based on the 13-billion parameter Meta Llama 2 model, this dataset encompasses 21466 medical discharge summaries extracted from the MIMIC-IV-Note dataset, with 156599 synthetically generated question-and-answer pairs, a subset of which was verified for accuracy by a physician. These pairs were generated by providing the model with a discharge summary and instructing it to generate question-and-answer pairs based on the contextual information present in the summaries. This work aims to generate data in support of the development of compact large language models capable of efficiently extracting information from medical notes and discharge summaries, thus enabling potential improvements for real-time decision-making processes in clinical settings. Additionally, accompanying the dataset is code facilitating question-and-answer pair generation from any medical and non-medical text. Despite the robustness of the presented dataset, it has certain limitations. The generation process was confined to a maximum context length of 6000 input tokens, owing to hardware constraints. The large language model's nature in generating these question-and-answer pairs may introduce an underlying bias or a lack in diversity and complexity. Future iterations should focus on rectifying these issues, possibly through diversified training and expanded verification procedures as well as the employment of more powerful large language models.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":81463293,"compressed_storage_size":17802417,"resource_type":"Database","access_policy":"Credentialed","topics":["mimic-iv","large language models","clinical question-answering","medical discharge summaries"],"source_url":"https://physionet.org/content/ehr-ds-qa/1.0.0/"},{"public_project_uuid":"6732ad34-9221-49ea-adf1-5d542c9edf09","slug":"nlp-opioid-behavior-detection","version":"1.0.0","core_doi":"10.13026/f0hv-7037","version_doi":"10.13026/qrje-z188","is_latest_version":true,"short_description":"Opioid-related aberrant behaviors (ORABs) detection Dataset (ODD) which is a large-size, expert-annotated, and multi-label classification benchmark dataset corresponding to the task","publish_date":"2024-01-11","title":"ODD: A Benchmark Dataset for the NLP-based Opioid Related Aberrant Behavior Detection","abstract":"Opioid related aberrant behaviors (ORAB) present novel risk factors for opioid overdose. Previously, ORAB have been mainly assessed by survey results and by monitoring drug administrations. Such methods however, cannot scale up and do not cover the entire spectrum of aberrant behaviors. On the other hand, ORAB are widely documented in electronic health record notes. This paper introduces a novel biomedical natural language processing benchmark dataset named ODD, for ORAB Detection Dataset. ODD is an expert-annotated dataset comprising of more than 750 publicly available EHR notes. ODD has been designed to identify ORAB from patients' EHR notes and classify them into nine categories; 1) Confirmed Aberrant Behavior, 2) Suggested Aberrant Behavior, 3) Opioids, 4) Indication, 5) Diagnosed opioid dependency, 6) Benzodiapines, 7) Medication Changes, 8) Central Nervous System-related, and 9) Social Determinants of Health.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":390164,"compressed_storage_size":108828,"resource_type":"Database","access_policy":"Credentialed","topics":["natural language processing","substance use","opioid related aberrant behavior"],"source_url":"https://physionet.org/content/nlp-opioid-behavior-detection/1.0.0/"},{"public_project_uuid":"ceefb9fb-f6ae-486b-837b-38fcb65404d7","slug":"perg-ioba-dataset","version":"1.0.0","core_doi":"10.13026/2t6a-xq52","version_doi":"10.13026/d24m-w054","is_latest_version":true,"short_description":"336 CSV records with 1354 PERG responses (microvolts) from 304 subjects at IOBA. Includes age (years), gender, diagnoses, and visual acuity in logMar scale.","publish_date":"2024-01-19","title":"A Comprehensive Dataset of Pattern Electroretinograms for Ocular Electrophysiology Research: The PERG-IOBA Dataset","abstract":"The pattern electroretinogram (PERG) is a valuable tool in ophthalmic electrophysiology, offering a non-invasive and objective method to evaluate central retinal function. By measuring electrical activity in the macula and retinal ganglion cells, PERG provides essential quantitative insights, especially when distinguishing between macular and optic nerve diseases remains challenging. This dataset comprises 1354 transient PERG responses from 304 subjects in 336 records, representing a comprehensive collection of ocular electrophysiological data. The evaluations took place at the Institute of Applied Ophthalmobiology (IOBA) at the University of Valladolid, Spain, from 2003 to 2022. Detailed clinical information, including age, gender, and diagnoses, is provided in CSV format. Additionally, the dataset, which guarantees the presence of at least one PERG signal for each eye, incorporates visual acuity measurements in logMar scale, facilitating assessments of vision quality. Addressing a significant gap in ocular electrophysiological signal repositories, this dataset serves as a valuable resource for the advancement of ophthalmology research. Researchers can explore a wide range of eye-related conditions and diseases, leading to improved diagnostic techniques, treatment strategies, and a deeper understanding of ocular electrophysiology.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":5977884,"compressed_storage_size":1002388,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/perg-ioba-dataset/1.0.0/"},{"public_project_uuid":"7ef1253d-32e8-46ba-bb5e-3156fcc40d1c","slug":"annotation-dataset-sdoh","version":"1.0.1","core_doi":"10.13026/p8zc-cs44","version_doi":"10.13026/zsgv-8w31","is_latest_version":true,"short_description":"Annotation dataset of social determinants of health from MIMC-III Clinical Care Database notes.","publish_date":"2024-01-25","title":"Annotation dataset of social determinants of health from MIMIC-III Clinical Care Database","abstract":"Social determinants of health (SDoH) have an important impact on patient outcomes but are incompletely collected from the electronic health records (EHR). This study researched the ability of large language models to extract SDoH from free text in EHRs, where they are most commonly documented, and explored the role of synthetic clinical text for improving the extraction of these scarcely documented, yet extremely valuable, clinical data. We developed annotation guidelines for sentence-level annotation of SDoH that are not reliably available as structured data in the EHR: employment, housing, transportation, parental status, relationship, and social support. Sentences were labeled for both the presence of an SDoH mention and the presence of an adverse SDoH mention. After finalizing the annotation guidelines, two annotators manually annotated a separate corpus, which cannot be released due to PHI. A total of 300/800 (37.5%) of these notes underwent dual annotation. Before adjudication, dually-annotated notes had a Krippendorf’s alpha agreement of 0.86 and Cohen’s Kappa of 0.86 for any SDoH mention categories. For adverse SDoH mentions, notes had a Krippendorf’s alpha agreement of 0.76 and Cohen’s Kappa of 0.76. As an external validation, 200 notes from MIMIC-III written by physicians, social workers, and nurses were manually annotated by a single annotator. Here, we release this manually annotated corpus of 200 MIMC-III notes.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1760080,"compressed_storage_size":530477,"resource_type":"Database","access_policy":"Credentialed","topics":["social determinants of health","natural language processing"],"source_url":"https://physionet.org/content/annotation-dataset-sdoh/1.0.1/"},{"public_project_uuid":"ed02ca90-aea8-4d72-bb3f-85e9f72840dc","slug":"rad-coreference-resolution","version":"1.0.0","core_doi":"10.13026/x1e4-5r04","version_doi":"10.13026/z67q-xy65","is_latest_version":true,"short_description":"RadCoref is a small subset of MIMIC-CXR with manually annotated coreference mentions and clusters. Based on the annotated data, we fine-tuned a deep neural model and used it to annotate the whole MIMIC-CXR dataset. Both data are available.","publish_date":"2024-01-30","title":"RadCoref: Fine-tuning coreference resolution for different styles of clinical narratives","abstract":"RadCoref is a small subset of MIMIC-CXR with manually annotated coreference mentions and clusters. The dataset is annotated by a panel of three cross-disciplinary experts with experience in clinical data processing following the i2b2 annotation scheme with minimum modification. The dataset consists of Findings and Impression sections extracted from full radiology reports. The dataset has 950, 25 and 200 section documents for training, validation, and testing, respectively. The training and validation sets are annotated by one annotator. The test set is annotated by two human annotators independently, of which the results are merged manually by the third annotator. The dataset aims to support the task of coreference resolution on radiology reports. Given that the MIMIC-CXR has been de-identified already, no protected health information (PHI) is included.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":358789869,"compressed_storage_size":256086653,"resource_type":"Database","access_policy":"Credentialed","topics":["natural language processing","radiology","coreference resolution"],"source_url":"https://physionet.org/content/rad-coreference-resolution/1.0.0/"},{"public_project_uuid":"4858a3b3-5735-4a9b-b21f-3c78f8c4c762","slug":"asclepius-r","version":"1.0.1","core_doi":"10.13026/1wf7-pm14","version_doi":"10.13026/s5rz-1j65","is_latest_version":false,"short_description":"Asclepius: Publicly Available Clinical Large Language Models with Synthetic Clinical Notes Asclepius-R: A instruction-finetuned large language model with MIMIC-III clinical notes","publish_date":"2024-01-30","title":"Asclepius-R : Clinical Large Language Model Built On MIMIC-III Discharge Summaries","abstract":"The development of large language models tailored for handling patients’ clinical notes is often hindered by the limited accessibility and usability of these notes due to strict privacy regulations. To address these challenges, we first create synthetic large-scale clinical notes using publicly available case reports extracted from biomedical literature. We then use these synthetic notes to train our specialized clinical large language model, Asclepius. While Asclepius is trained on synthetic data, we assess its potential performance in real-world applications by evaluating it using real clinical notes. We benchmark Asclepius against several other large language models, including GPT-3.5-turbo and other open-source alternatives. To further validate our approach using synthetic notes, we also compare Asclepius with its variants trained on real clinical notes. Our findings convincingly demonstrate that synthetic clinical notes can serve as viable substitutes for real ones when constructing high-performing clinical language models. This conclusion is supported by detailed evaluations conducted by both GPT-4 and medical professionals. All resources—including weights, codes, and data—used in the development of Asclepius are made publicly accessible for future research. Specifically, this repository contains Asclepius-R, a variant of Asclepius that was trained on MIMIC-III discharge summaries. All other resource are also publicly accessible.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":79058055436,"compressed_storage_size":73163600170,"resource_type":"Model","access_policy":"Credentialed","topics":["synthetic notes","large language model","asclepius","synthetic clinical notes","llm","open-source","clinical notes","clinical llm"],"source_url":"https://physionet.org/content/asclepius-r/1.0.1/"},{"public_project_uuid":"2bd9bebf-4e59-40c6-93c4-e282cec501f4","slug":"multimodal-satellite-data","version":"1.0.0","core_doi":"10.13026/s5kt-3p13","version_doi":"10.13026/xr5s-xe24","is_latest_version":true,"short_description":"Multi-Modal Satellite imagery Dataset in Colombia: A public health analysis with spatiotemporally aligned satellite images and its corresponding metadata across 81 municipalities (2016-2018), facilitating multimodal AI applications.","publish_date":"2024-01-30","title":"A Multi-Modal Satellite Imagery Dataset for Public Health Analysis in Colombia","abstract":"We introduce a cost-effective public health analysis solution for low- and middle-income countries—the Multi-Modal Satellite Imagery Dataset in Colombia. By leveraging high-quality, spatiotemporally aligned satellite images and corresponding metadata, the dataset integrates economic, demographic, meteorological, and epidemiological data. Employing a single forwards and a forward-backward technique ensures clear satellite images with minimal cloud cover for every epi-week, significantly enhancing overall data quality. The extraction process utilizes the satellite extractor package powered by the SentinelHub API, resulting in a comprehensive dataset of 12,636 satellite images from 81 municipalities in Colombia between 2016 and 2018, along with relevant metadata. Beyond expediting public health data analysis across diverse locations and timeframes, this versatile framework consistently captures multimodal features. Its applications extend to various realms in multimodal AI, encompassing deforestation monitoring, forecasting education indices, water quality assessment, tracking extreme climatic events, addressing epidemic illnesses, and optimizing precision agriculture.
","license":{"name":"Creative Commons Zero 1.0 Universal Public Domain Dedication"},"dua":null,"main_storage_size":69834315241,"compressed_storage_size":69127501075,"resource_type":"Database","access_policy":"Open","topics":["satellite imagery","multimodality"],"source_url":"https://physionet.org/content/multimodal-satellite-data/1.0.0/"},{"public_project_uuid":"6d50eccf-7c9f-40e0-8507-a46b4b57223a","slug":"curated-oncology-reports","version":"1.0","core_doi":"10.13026/vwq3-fj85","version_doi":"10.13026/v69y-xa45","is_latest_version":true,"short_description":"Medical oncology progress notes annotated with advanced, comprehensive oncology-relevant concepts and relationships.","publish_date":"2024-02-07","title":"CORAL: expert-Curated medical Oncology Reports to Advance Language model inference","abstract":"Both medical care and observational studies in oncology require a thorough understanding of a patient's disease progression and treatment history, often elaborately documented within clinical notes. As large language models (LLMs) are becoming more popular, it becomes important to evaluate their potential in oncology. However, no current information representation schema fully encapsulates the diversity of oncology information within clinical notes, and no comprehensively annotated oncology notes exist publicly, thereby limiting a thorough evaluation. We curated a new fine-grained, expert-labeled dataset of 40 de-identified breast and pancreatic cancer progress notes at University of California, San Francisco, and assessed three recent LLMs (GPT-4, GPT-3.5-turbo, and FLAN-UL2) in zero-shot extraction of detailed oncological information from two narrative sections of clinical progress notes. Model performance was quantified with BLEU-4, ROUGE-1, and exact match (EM) F1-score evaluation metrics. Our team of oncology fellows and medical students manually annotated 9028 entities, 9986 modifiers, and 5312 relationships. The GPT-4 model exhibited overall best performance, with an average BLEU score of 0.73, an average ROUGE score of 0.72, an average EM-F1-score of 0.51, and an average accuracy of 68% (expert manual evaluation on 20 notes). GPT-4 was proficient in tumor characteristics and medication extraction, and demonstrated superior performance in inferring symptoms due to cancer and considerations of future medications. Common errors included partial responses with missing information and hallucinations with note-specific information. LLMs are promising for performing reliable information extraction for clinical research, complex population management, and documenting quality patient care, but there is a need for further improvements.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":22867408,"compressed_storage_size":2780829,"resource_type":"Database","access_policy":"Credentialed","topics":["oncology","natural language processing","information extraction","artificial intelligence","large language models","electronic health records"],"source_url":"https://physionet.org/content/curated-oncology-reports/1.0/"},{"public_project_uuid":"3455d73e-1a38-42b5-b101-b488760ad768","slug":"heart-rate-fragmentation-code","version":"1.0.0","core_doi":"10.13026/krqe-pj80","version_doi":"10.13026/0mzj-gn98","is_latest_version":true,"short_description":"Heart rate fragmentation: a new method for the analysis of cardiac interbeat interval time series. The code provided can be run in Windows, Mac and Linux machines.","publish_date":"2024-02-14","title":"Software for computing Heart Rate Fragmentation","abstract":"Heart rate fragmentation (HRF) is a new method for assessing neuroautonomic integrity based on the analysis of short-term (high-frequency [HF]) heart rate dynamics. The code (in AWK) provided here is for the computation of three different metrics, PIP, PNNLS, and PNNSS introduced in 2017 by Costa and colleagues.
\n\nThe motivation for developing HRF was the recognition that changes in heart rate (HR) above the respiratory frequency not only inflate the value of traditional heart rate variability (HRV) metrics, such as rMSSD and HF power, but are a dynamical marker of cardioautonomic dysfunction.
\n\nIn healthy subjects, HR increases with inspiration and decreases with expiration, defining a “high-frequency” pattern referred to as respiratory sinus arrhythmia. This physiologic oscillation is driven primarily by changes in vagal tone. In contexts in which cardio-respiratory coupling is degraded, e.g., with aging and disease, non-vagally mediated oscillations whose frequencies are higher than the breathing frequency may develop. Non-vagally mediated fluctuations are characterized by an excess of reversals in HR acceleration. The simplest measure of HRF is the percentage of “inflection points,” (PIP) i.e., of changes in HR acceleration sign. Of key importance is fact that traditional short-term HRV measures can be high for two opposite reasons, one healthy – respiratory sinus arrhythmia – consistent with high vagal tone modulation, and one pathologic – fragmented rhythms – consistent with decreased/impaired vagal activity.
\n\nIncreased HRF has been shown to be independently associated with higher risk of incident major adverse cardiovascular events and atrial fibrillation. These results have been independently confirmed, as described in Background section.
","license":{"name":"GNU General Public License version 3"},"dua":null,"main_storage_size":2619213,"compressed_storage_size":2307032,"resource_type":"Software","access_policy":"Open","topics":["cardiovascular disease","vagal tone","time series analysis","prediction of atrial fibrillation","cardiac autonomic function","prediction of cardiovascular events","prediction of cognitive decline","heart rate variability","heart rate fragmentation","aging"],"source_url":"https://physionet.org/content/heart-rate-fragmentation-code/1.0.0/"},{"public_project_uuid":"b0cfea51-1224-4c27-a185-643bdd9a4e4e","slug":"mimic-iv-fhir","version":"1.0","core_doi":"10.13026/9hhp-ps57","version_doi":"10.13026/cqt2-0b27","is_latest_version":false,"short_description":"MIMIC-IV and MIMIC-IV-ED data mapped into FHIR resources.","publish_date":"2024-02-20","title":"MIMIC-IV on FHIR","abstract":"Fast Healthcare Interoperability Resources (FHIR) has emerged as a robust standard for healthcare data exchange. To explore the use of FHIR for the process of data harmonization, we converted the Medical Information Mart for Intensive Care IV (MIMIC-IV) and MIMIC-IV Emergency Department (MIMIC-IV-ED) databases into FHIR. We extended base FHIR to encode information in MIMIC-IV and aimed to retain the data in FHIR with minimal additional processing, aligning to US Core v4.0.0 where possible. A total of 24 profiles were created for MIMIC-IV data, and an additional 6 profiles were created for MIMIC-IV-ED data. All MIMIC terminology was converted into code systems and value sets, as necessary. We hope MIMIC-IV in FHIR provides a useful restructuring of the data to support applications around data harmonization, interoperability, and other areas of research.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":31229847916,"compressed_storage_size":30977632213,"resource_type":"Database","access_policy":"Credentialed","topics":["fhir","electronic health record","mimic-iv"],"source_url":"https://physionet.org/content/mimic-iv-fhir/1.0/"},{"public_project_uuid":"26933698-3c38-4b2d-8491-31f3feaeddba","slug":"corpus-fungal-infections","version":"1.0.2","core_doi":"10.13026/g3t3-2c54","version_doi":"10.13026/m1rk-ns13","is_latest_version":true,"short_description":"A corpus of cytology and histopathology reports annotated for terminology relevant to fungal infections. Ideal for validation of named entity recognition and relation extraction methods.","publish_date":"2024-02-20","title":"CHIFIR: Cytology and Histopathology Invasive Fungal Infection Reports","abstract":"Surveillance of invasive fungal infection (IFI) in clinical settings is a laborious process requiring a detailed review of patient medical history. One of the key sources of clinical information is cytology and histopathology reports: pathologist-produced free-text reports outlining the macroscopic and microscopic structure of a specimen. The data was generated to facilitate the development of an automated tool for the detection of IFI. The Cytology and Histopathology Invasive Fungal Infection Reports (CHIFIR) corpus contains 283 de-identified reports annotated by infectious diseases physicians for terminology relevant to the IFI diagnosis. These include IFI-specific concepts and certainty cues, as well as relations between them. We release the annotation schema and the original reports along with the corresponding annotation files. We anticipate the CHIFIR corpus to be useful in the development and validation of named entity recognition and relation extraction methods with a focus on clinical data. Such methods can be instrumental in processing other types of clinical documentation (radiology reports, clinical notes, nursing notes) with various downstream tasks in mind.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":504944,"compressed_storage_size":401531,"resource_type":"Database","access_policy":"Credentialed","topics":["information extraction","nlp","clinical documentation","invasive fungal infections"],"source_url":"https://physionet.org/content/corpus-fungal-infections/1.0.2/"},{"public_project_uuid":"c82d8948-31ee-4199-b117-423eeeadb1f2","slug":"echo-note-to-num","version":"1.0.0","core_doi":"10.13026/0myn-kk52","version_doi":"10.13026/xhrz-ht59","is_latest_version":true,"short_description":"A structured echocardiogram database derived from 43,472 observational notes obtained during echocardiogram studies conducted in the intensive care unit at the Beth Israel Deaconess Medical Center between 2001 and 2012.","publish_date":"2024-02-24","title":"EchoNotes Structured Database derived from MIMIC-III (ECHO-NOTE2NUM)","abstract":"The EchoNotes Structured Database derived from MIMIC-III (ECHO-NOTE2NUM) is a structured echocardiogram database derived from 43,472 observational notes obtained during echocardiogram studies conducted in the intensive care unit at the Beth Israel Deaconess Medical Center between 2001 and 2012. The database encompasses various aspects of cardiac structure and function, including cavity size, wall thickness, systolic and diastolic function, valve regurgitation and stenosis, as well as pulmonary pressures. To facilitate extensive data analysis, the clinical notes were transformed into a structured numerical format. Within each echocardiogram report sentence, specific words or phrases were identified to describe abnormal findings, and a severity staging system using numeric categories was established. This large publicly-accessible database of structured echocardiogram data holds significant potential as a tool to investigate cardiovascular disease in the intensive care unit.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":115734885,"compressed_storage_size":17868443,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/echo-note-to-num/1.0.0/"},{"public_project_uuid":"fe1970c2-436a-4d53-a7ac-7b0b0ded0a80","slug":"openox-repo","version":"1.0.0","core_doi":"10.13026/e821-1858","version_doi":"10.13026/cc78-ad74","is_latest_version":false,"short_description":"A repository of matched arterial oxygen and pulse oximeter readings obtained under controlled conditions, with high-frequency physiologic waveforms and skin color measurements.","publish_date":"2024-02-27","title":"OpenOximetry Repository","abstract":"The OpenOximetry Repository is a structured database designed to store clinical and laboratory pulse oximetry data and allows for consolidation of data sets held by collaborating organizations. Matched or independent readings of oxygen saturations, arterial blood gas samples, high frequency processed and unprocessed waveforms, and other physiologic metadata are supported (e.g. skin color readings). Contribution of data to the repository is encouraged; a data dictionary and standardized data collection protocols are provided to ensure consistent data archival and interpretation. The goal of this dataset is to enable researchers to interrogate the effects of physiologic and device variables on pulse oximeter performance.
\n\nA set of controlled desaturation studies is included in the initial release with ongoing updates planned as more data are available from both controlled lab desaturation studies and real world prospective clinical trials. Data captured during these studies may include paired pulse oximeter and arterial blood gas readings throughout a range of oxygen saturations, patient demographics and comorbidities, longitudinal quantitative and qualitative skin color measurement (using multiple methods at multiple anatomic locations), finger diameter, continuous ECG, arterial blood pressure, EtCO2/EtO2, unprocessed photoplethysmography (PPG), pH, PaCO2, hemoglobin, met-Hgb, CO-Hgb, respiratory rate and multiple other variables depending on the data source
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":2533046892,"compressed_storage_size":1584371188,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/openox-repo/1.0.0/"},{"public_project_uuid":"fd483ce8-4394-4bee-bff4-896948149df3","slug":"chexmask-cxr-segmentation-data","version":"0.4","core_doi":"10.13026/6eky-y831","version_doi":"10.13026/pgag-by42","is_latest_version":false,"short_description":"CheXmask Database is a 657,566 uniformly annotated chest radiographs with segmentation masks. Images were segmented using HybridGNet, with automatic quality control indicated by RCA scores.","publish_date":"2024-03-01","title":"CheXmask Database: a large-scale dataset of anatomical segmentation masks for chest x-ray images","abstract":"The CheXmask Database presents a comprehensive, uniformly annotated collection of chest radiographs, constructed from five public databases: ChestX-ray8, Chexpert, MIMIC-CXR-JPG, Padchest and VinDr-CXR. The database aggregates 657,566 anatomical segmentation masks derived from images which have been processed using the HybridGNet model to ensure consistent, high-quality segmentation. To confirm the quality of the segmentations, we include in this database individual Reverse Classification Accuracy (RCA) scores for each of the segmentation masks. This dataset is intended to catalyze further innovation and refinement in the field of semantic chest X-ray analysis, offering a significant resource for researchers in the medical imaging domain.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":40089505711,"compressed_storage_size":16784459188,"resource_type":"Database","access_policy":"Open","topics":["chest x-ray segmentation","medical image segmentation","automatic quality assesment"],"source_url":"https://physionet.org/content/chexmask-cxr-segmentation-data/0.4/"},{"public_project_uuid":"a93d4c00-0273-4cc5-bd55-d90655db94a1","slug":"discharge-me","version":"1.2","core_doi":"10.13026/27pt-1259","version_doi":"10.13026/4a0k-4360","is_latest_version":false,"short_description":"Data for the \"Discharge Me!\" Shared Task on Streamlining Discharge Documentation for BioNLP ACL'24","publish_date":"2024-03-05","title":"Discharge Me: BioNLP ACL'24 Shared Task on Streamlining Discharge Documentation","abstract":""Discharge Me!", part of the BioNLP workshop co-located with ACL 2024, seeks to alleviate the significant burden on clinicians who dedicate substantial time to crafting detailed discharge notes in the EHR. Participants in the task will explore approaches to generating "Brief Hospital Course" and "Discharge Instructions" sections of the discharge summary using a subset of MIMIC-IV-Note and MIMIC-IV-ED that have been compiled by the task organizers. The full dataset (comprised of a defined training, validation, phase 1 testing, and phase 2 testing sets) consists of 109,168 emergency department admissions. The competition is being hosted on the Codabench platform, which will manage team registration, results submission, and score evaluation.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":658325628,"compressed_storage_size":658428949,"resource_type":"Challenge","access_policy":"Credentialed","topics":["generation","discharge summary","bionlp","acl"],"source_url":"https://physionet.org/content/discharge-me/1.2/"},{"public_project_uuid":"c07d5a11-87fb-4f2d-b011-0b76022b37a6","slug":"mimic-cxr-jpg","version":"2.1.0","core_doi":"10.13026/th9c-ae10","version_doi":"10.13026/jsn5-t979","is_latest_version":true,"short_description":"Chest x-rays in JPG format with structured labels derived from the associated radiology report.","publish_date":"2024-03-12","title":"MIMIC-CXR-JPG - chest radiographs with structured labels","abstract":"The MIMIC Chest X-ray JPG (MIMIC-CXR-JPG) Database v2.0.0 is a large publicly available dataset of chest radiographs in JPG format with structured labels derived from free-text radiology reports. The MIMIC-CXR-JPG dataset is wholly derived from MIMIC-CXR, providing JPG format files derived from the DICOM images and structured labels derived from the free-text reports. The aim of MIMIC-CXR-JPG is to provide a convenient processed version of MIMIC-CXR, as well as to provide a standard reference for data splits and image labels. The dataset contains 377,110 JPG format images and structured labels derived from the 227,827 free-text radiology reports associated with these images. The dataset is de-identified to satisfy the US Health Insurance Portability and Accountability Act of 1996 (HIPAA) Safe Harbor requirements. Protected health information (PHI) has been removed. The dataset is intended to support a wide body of research in medicine including image understanding, natural language processing, and decision support.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":612352535731,"compressed_storage_size":608922370063,"resource_type":"Database","access_policy":"Credentialed","topics":["computer vision","chest x-ray","mimic","deep learning","radiology"],"source_url":"https://physionet.org/content/mimic-cxr-jpg/2.1.0/"},{"public_project_uuid":"f1c9de66-1b69-450b-9162-2b2f92a7ed10","slug":"chexchonet","version":"1.0.0","core_doi":"10.13026/zsej-wt49","version_doi":"10.13026/kp08-ws25","is_latest_version":true,"short_description":"Early detection of heart failure is vital for improving outcomes. The dataset contains 71,589 CXRs paired with gold standard labels from echocardiograms to enable the training of models to detect pathologies indicative of early stage heart failure.","publish_date":"2024-03-20","title":"CheXchoNet: A Chest Radiograph Dataset with Gold Standard Echocardiography Labels","abstract":"Existing chest radiograph datasets, such as CheXpert and ChestX-ray14, have driven the development of new machine learning approaches to achieve expert or near-expert level performance on a variety of tasks. The primary focus of models developed using these datasets has been to replicate human-level performance by training on labels computationally extracted from radiology reports. We propose a different paradigm: pair an existing diagnostic test with labels from a more accurate, higher fidelity diagnostic test. This approach seeks to ask whether data from a cheaper, lower fidelity diagnostic test contains information for detection of pathologies using more accurate, gold standard labels. In the context of chest X-rays, a good example is the radiologic comment of cardiomegaly, a catch-all term or an abnormally enlarged heart. Cardiomegaly is known to be poorly predictive of cardiac disease and does not trigger meaningful clinical action. Instead, we can pair chest X-rays with gold standard structural heart disease labels derived from echocardiograms conducted on the same patients. This resource contains 71,589 unique chest X-rays from 24,689 different patients paired with key echocardiography measurements indicative of left ventricular hypertrophy and dilated left ventricle, pathologies which occur during early stage heart failure. The data also includes information about the relative times of the chest X-rays, the age/sex of the patient at the time of recording, and related metadata information. This data can be used as a resource for the community to build novel approaches to detect clinically actionable labels.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":2916338326,"compressed_storage_size":2905808151,"resource_type":"Database","access_policy":"Restricted","topics":["early detection","heart failure","cardiac structural abnormalties","deep learning","chest x-rays"],"source_url":"https://physionet.org/content/chexchonet/1.0.0/"},{"public_project_uuid":"e525aef9-cdc8-4bb3-b3b0-6fc8ca0a0dcd","slug":"respiratory-heartrate-dataset","version":"1.0.0","core_doi":"10.13026/cpeg-7b46","version_doi":"10.13026/e4dt-f689","is_latest_version":true,"short_description":"Respiratory and cardiovascular data collected from 20 subjects. Pressure, flow, aeration, and heart-rate data were collected during trials which included resting breathing, CPAP at varied PEEP settings, breath-holds, and forced expiratory manoeuvres.","publish_date":"2024-03-20","title":"Respiratory and heart rate monitoring dataset from aeration study","abstract":"A study was conducted to collect respiratory pressure and flow data for model-based assessment, alongside electrical impedance tomography (EIT) aeration, electrocardiogram (ECG), and heart-rate belt (HRB) data. A 20 subjects set was selected with no known significant respiratory abnormalities, under University of Canterbury Human Research Ethics Committee (HREC) consent (HREC 2023/30/LR-PS). Subject demographic data was collected using a combination of measurements and a self-reported questionnaire. Subject demographic information to ascertain: sex; height; weight; age; any history of asthma, smoking, or vaping; and approximate thoracic cage and breast tissue volumes. The trial was setup with subjects breathing through the pressure and flowmeter using a full-face mask and filter. Inspiratory and expiratory flow were separated using one-way valves for measurement purposes and both ports were rapidly occluded every 200ms to generate passive mechanics assessment intervals. A continuous positive airway pressure (CPAP) machine was used to provide positive airway pressure (PEEP) in series with the inspiratory port. Three trials were conducted per subject. Two of which, involved increasing PEEP (0, 4, and 8 cmH2O), and in one of which subjects were asked to perform two 10s breath holds at each level. The third was a recording of three forced expiratory manoeuvres (FEM). These tests were conducted to provide an initial comparison of current respiratory techniques to model-based methods, which aims to inform the design of future clinical testing on subjects with known respiratory abnormalities.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":5659050955,"compressed_storage_size":3329463131,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/respiratory-heartrate-dataset/1.0.0/"},{"public_project_uuid":"f6c0eb9b-ae8c-4dd9-90e2-fd40b9e96850","slug":"asclepius-r","version":"1.1.0","core_doi":"10.13026/1wf7-pm14","version_doi":"10.13026/ckbf-pt70","is_latest_version":true,"short_description":"Asclepius: Publicly Available Clinical Large Language Models with Synthetic Clinical Notes Asclepius-R: A instruction-finetuned large language model with MIMIC-III clinical notes","publish_date":"2024-03-26","title":"Asclepius-R : Clinical Large Language Model Built On MIMIC-III Discharge Summaries","abstract":"The development of large language models tailored for handling patients’ clinical notes is often hindered by the limited accessibility and usability of these notes due to strict privacy regulations. To address these challenges, we first create synthetic large-scale clinical notes using publicly available case reports extracted from biomedical literature. We then use these synthetic notes to train our specialized clinical large language model, Asclepius. While Asclepius is trained on synthetic data, we assess its potential performance in real-world applications by evaluating it using real clinical notes. We benchmark Asclepius against several other large language models, including GPT-3.5-turbo and other open-source alternatives. To further validate our approach using synthetic notes, we also compare Asclepius with its variants trained on real clinical notes. Our findings convincingly demonstrate that synthetic clinical notes can serve as viable substitutes for real ones when constructing high-performing clinical language models. This conclusion is supported by detailed evaluations conducted by both GPT-4 and medical professionals. All resources—including weights, codes, and data—used in the development of Asclepius are made publicly accessible for future research. Specifically, this repository contains Asclepius-R, a variant of Asclepius that was trained on MIMIC-III discharge summaries. All other resource are also publicly accessible.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":158076664927,"compressed_storage_size":146315016508,"resource_type":"Model","access_policy":"Credentialed","topics":["large language model","synthetic clinical notes","synthetic notes","asclepius","open-source","llm","clinical llm","clinical notes"],"source_url":"https://physionet.org/content/asclepius-r/1.1.0/"},{"public_project_uuid":"150be842-18f1-4a1c-b911-42d6ea8e6e58","slug":"scientisst-move-biosignals","version":"1.0.1","core_doi":"10.13026/0ppk-ha30","version_doi":"10.13026/hyxq-r919","is_latest_version":true,"short_description":"Multimodal (ECG, EMG, EDA, PPG, TEMP, ACC) biosignal dataset of everyday activities. Created with 3 wearable devices based on ScientISST Sense and Empatica E4.","publish_date":"2024-03-26","title":"ScientISST MOVE: Annotated Wearable Multimodal Biosignals recorded during Everyday Life Activities in Naturalistic Environments","abstract":"Existing datasets containing physiological data are mostly acquired at rest or in controlled scenarios. As a result, algorithms developed using such data may not perform as well as with biosignals acquired in dynamic and uncontrolled environments. ScientiSST MOVE is a multimodal dataset recording natural everyday activities, including lifting a chair, greeting, gesticulating, walking and running. Data was collected using three wearable devices, namely: a chestband, an armband, and the Empatica E4 wristband. This setup enabled recording of multi-channel Electrodermal Activity (EDA), Photoplethysmography (PPG) and Electrocardiography (ECG). Additionally, recordings were also made for bicep Electromyography (EMG), wrist temperature and chest and wrist actigraphy. A total of 17 healthy volunteers participated in the experimental data acquisition sessions, resulting in an average of 37 useful minutes of synchronised data from all sensors. ScientISST MOVE has been primarily designed to study the effect of daily activities on physiological data acquisition. Having been acquired with multiple wearable devices, some of which measuring the same modalities, it can also be useful in signal quality comparison studies.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":468609476,"compressed_storage_size":206596792,"resource_type":"Database","access_policy":"Open","topics":["greet","wearable","multimodal","lift","uncontrolled environments","run","jump","gesticulate","walk"],"source_url":"https://physionet.org/content/scientisst-move-biosignals/1.0.1/"},{"public_project_uuid":"3270f2fa-29f8-4fcf-8f3d-e6355fb882a1","slug":"parkinsons-disease-smartwatch","version":"1.0.0","core_doi":"10.13026/paqx-kc69","version_doi":"10.13026/m0w9-zx22","is_latest_version":true,"short_description":"The PADS dataset contains smartwatch-based records from interactive neurological assessments of Parkinsons disease patients, differential diagnoses and healthy controls. The data is complemented with non-motor symptoms and medical history information","publish_date":"2024-03-26","title":"PADS - Parkinsons Disease Smartwatch dataset","abstract":"Parkinson’s disease (PD) is the second-most common neurodegenerative disorder, while incidence and worldwide burden are further increasing. In the era of digital health transformation, smart devices and mobile sensors, including smartphones and smartwatches, can provide an affordable source to capture and analyse digital objective biomarkers. These can aid in early diagnosis and studying phenotypical characteristics. The Parkinson’s Disease Smartwatch (PADS) dataset comprises clinical assessments of a broad spectrum of PD patients, other similar movement disorders and healthy controls. The assessments were recorded using a smart-device-based system consisting of two smartwatches and one smartphone. The two smartwatches were worn on each of the patient’s wrists and their sensors synchronously recorded 11 interactive movement tasks that were designed by expert neurologists to provoke subtle changes in movement pathologies. In total, 5159 measurement steps of 469 individuals were captured. The PADS dataset includes all acceleration and rotation sensor signals, as well as details on movement steps, demographics, medical history, and PD-specific non-motor symptoms. We believe that our extensively annotated dataset provides a well-suited data foundation for the training, validation, and optimisation of future technology- and sensor-driven systems for movement disorders.
","license":{"name":"Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":1517701375,"compressed_storage_size":770744328,"resource_type":"Database","access_policy":"Open","topics":["movement disorders","wearables","parkinsons disease"],"source_url":"https://physionet.org/content/parkinsons-disease-smartwatch/1.0.0/"},{"public_project_uuid":"44e41ec0-6557-411b-8c8a-2471f637b8f5","slug":"radialog-instruct-dataset","version":"1.0.0","core_doi":"10.13026/pxb3-eh70","version_doi":"10.13026/zecj-bh52","is_latest_version":false,"short_description":"Image-based instruct data for Chest X-Ray understanding and analysis.","publish_date":"2024-03-26","title":"RaDialog Instruct Dataset","abstract":"Conversational AI tools that can generate and discuss clinically correct radiology reports for a given medical image have the potential to transform radiology. Such a human-in-the-loop radiology assistant could facilitate a collaborative diagnostic process, thus saving time and improving the quality of reports. Towards this goal, we introduce RaDialog, the first thoroughly evaluated and publicly available large vision-language model for radiology report generation and interactive dialog. To keep the conversational abilities of the underlying LLM, we propose a comprehensive, semi-automatically labeled, image-grounded instruct dataset for chest X-ray radiology tasks. The dataset includes a variety of tasks, such as report correction, summarization or finding prediction. By training with this dataset, our method achieves state-of-the-art clinical correctness in report generation and shows impressive abilities in interactive tasks such as correcting reports and answering questions, serving as a foundational step toward clinical dialog systems.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":795152254,"compressed_storage_size":106838136,"resource_type":"Database","access_policy":"Credentialed","topics":["radiology report generation","large vision-language models","medical image understaning","radiology assistant","radiology chatbot"],"source_url":"https://physionet.org/content/radialog-instruct-dataset/1.0.0/"},{"public_project_uuid":"3e9f0158-472a-4151-9faf-470b0440e5d8","slug":"ehr-notes-qa-llms","version":"1.0.0","core_doi":"10.13026/mjgm-x360","version_doi":"10.13026/kvca-f224","is_latest_version":false,"short_description":"A patient-specific question answering benchmark tailored for evaluating Large Language Models (LLMs) in clinical environments","publish_date":"2024-04-03","title":"EHRNoteQA: A Patient-Specific Question Answering Benchmark for Evaluating Large Language Models in Clinical Settings","abstract":"We introduce EHRNoteQA, a patient-specific question answering benchmark tailored for evaluating Large Language Models (LLMs) in clinical environments. Based on MIMIC-IV Electronic Health Record (EHR), a team of three medical professionals has curated the dataset comprising 962 unique questions, each linked to a specific patient's EHR clinical notes. Our comprehensive evaluation on various large language models showed that their scores on EHRNoteQA correlate more closely with their performance in addressing real-world medical questions evaluated by clinicians than their scores from other LLM benchmarks. This emphasizes the importance of EHRNoteQA in assessing Large Language Models (LLMs) for medical purposes and underscores its contribution to incorporating LLMs into healthcare infrastructures.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":806317,"compressed_storage_size":192076,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/ehr-notes-qa-llms/1.0.0/"},{"public_project_uuid":"ecd3f411-d7b9-4ce6-acef-fdf94563b16d","slug":"discharge-me","version":"1.3","core_doi":"10.13026/27pt-1259","version_doi":"10.13026/0zf5-fx50","is_latest_version":true,"short_description":"Data for the \"Discharge Me!\" Shared Task on Streamlining Discharge Documentation for BioNLP ACL'24","publish_date":"2024-04-12","title":"Discharge Me: BioNLP ACL'24 Shared Task on Streamlining Discharge Documentation","abstract":""Discharge Me!", part of the BioNLP workshop co-located with ACL 2024, seeks to alleviate the significant burden on clinicians who dedicate substantial time to crafting detailed discharge notes in the EHR. Participants in the task will explore approaches to generating "Brief Hospital Course" and "Discharge Instructions" sections of the discharge summary using a subset of MIMIC-IV-Note and MIMIC-IV-ED that have been compiled by the task organizers. The full dataset (comprised of a defined training, validation, phase 1 testing, and phase 2 testing sets) consists of 109,168 emergency department admissions. The competition is being hosted on the Codabench platform, which will manage team registration, results submission, and score evaluation.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":732574323,"compressed_storage_size":732690233,"resource_type":"Challenge","access_policy":"Credentialed","topics":["acl","generation","discharge summary","bionlp"],"source_url":"https://physionet.org/content/discharge-me/1.3/"},{"public_project_uuid":"3acb8ecf-4fcc-4ec0-b472-6c89fdde0237","slug":"carmen-i","version":"1.0.1","core_doi":"10.13026/bkwd-3j50","version_doi":"10.13026/x7ed-9r91","is_latest_version":true,"short_description":"CARMEN-I is a Spanish corpus of 2,000 clinical records from Hospital Clínic, Barcelona. It covers COVID-19 patients and comorbidities, serving as a resource for training clinical NLP models and researchers in NLP applied to clinical documents.","publish_date":"2024-04-20","title":"CARMEN-I: A resource of anonymized electronic health records in Spanish and Catalan for training and testing NLP tools","abstract":"The CARMEN-I corpus comprises 2,000 clinical records, encompassing discharge letters, referrals, and radiology reports from Hospital Clínic of Barcelona between March 2020 and March 2022. These reports, primarily in Spanish with some Catalan sections, cover COVID-19 patients with diverse comorbidities like kidney failure, cardiovascular diseases, malignancies, and immunosuppression. The corpus underwent thorough anonymization, validation, and expert annotation, replacing sensitive data with synthetic equivalents. A subset of the corpus features annotations of medical concepts by specialists, encompassing symptoms, diseases, procedures, medications, species, and humans (including family members). CARMEN-I serves as a valuable resource for training and assessing clinical NLP techniques and language models, aiding tasks like de-identification, concept detection, linguistic modifier extraction, document classification, and more. It also facilitates training researchers in clinical NLP and is a collaborative effort involving Barcelona Supercomputing Center's NLP4BIA team, Hospital Clínic, and Universitat de Barcelona's CLiC group.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":23681754,"compressed_storage_size":16379031,"resource_type":"Database","access_policy":"Contributor Review","topics":["de-identification","clinical ner","anonymization"],"source_url":"https://physionet.org/content/carmen-i/1.0.1/"},{"public_project_uuid":"4254eba0-45fa-4074-a971-3b1528e3f7e1","slug":"ann-pt-summ","version":"1.0.0","core_doi":"10.13026/m6hf-dq94","version_doi":"10.13026/a66y-aa53","is_latest_version":false,"short_description":"Annotations for unsupported facts in 100 original MIMIC patient summaries (discharge instructions) and hallucinations in 100 Large Language Model (LLM) generated patient summaries labeled by two medical experts.","publish_date":"2024-04-28","title":"Medical Expert Annotations of Unsupported Facts in Doctor-Written and LLM-Generated Patient Summaries","abstract":"Large language models in healthcare can generate informative patient summaries while reducing the documentation workload of healthcare professionals. However, these models are prone to producing hallucinations, that is, generating unsupported information, which is problematic in the sensitive healthcare domain. To better characterize unsupported facts in medical texts, we developed a rigorous labeling protocol. Following this protocol, two medical experts annotated unsupported facts in 100 doctor-written summaries from the MIMIC-IV-Note Discharge Instructions and hallucinations 100 LLM-generated patient summaries. Here, we are releasing two datasets based on these annotations: Hallucinations-MIMIC-DI and Hallucinations-Generated-DI. We find that using these datasets to train on hallucination-free examples effectively reduces hallucinations for both Llama 2 (2.60 to 1.55 hallucinations per summary) and GPT-4 (0.70 to 0.40). Furthermore, we created a preprocessed version of the MIMIC-IV-Notes Discharge Instructions, releasing both a full-context version (MIMIC-IV-Note-Ext-DI) and a version that only uses the Brief Hospital Course for context (MIMIC-IV-Note-Ext-DI-BHC).
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":3292398748,"compressed_storage_size":1131368148,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/ann-pt-summ/1.0.0/"},{"public_project_uuid":"de77bb2e-74e7-4445-97ff-38955a97e1d4","slug":"dreamt","version":"1.0.0","core_doi":"10.13026/dztc-dv77","version_doi":"10.13026/62an-cb28","is_latest_version":false,"short_description":"Dataset for Real-time sleep stage EstimAtion using Multisensor wearable Technology","publish_date":"2024-04-30","title":"DREAMT: Dataset for Real-time sleep stage EstimAtion using Multisensor wearable Technology","abstract":"Sleep is an intrinsic part of human life, and recent advancements in wearable technology and machine learning have promised continuous and non-invasive methods of tracking sleep health and patterns, providing an important facet to a more holistic understanding of well-being. However, it is still challenging to achieve consistent and reliable real-time estimates of sleep stages using only smartwatches. This is especially true for individuals with irregular sleep patterns or sleep disorders. A major contributing factor is the distinct lack of publicly accessible, large-scale datasets that allow researchers and engineers to validate their wearable sleep staging algorithms against a population with diverse sleep patterns. Here, we present DREAMT, Dataset for Real-time sleep stage EstimAtion using Multisensor wearable Technology, a new dataset collected from 100 participants, which includes high-resolution signals from a smartwatch, expert sleep technician-annotated sleep stage labels, and clinical metadata related to sleep health and disorders.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":14108970053,"compressed_storage_size":1422765357,"resource_type":"Database","access_policy":"Restricted","topics":["wearable","biomedical","sleep disorders","time series classification"],"source_url":"https://physionet.org/content/dreamt/1.0.0/"},{"public_project_uuid":"7270bed4-d664-45e7-bd80-24c43cd0be4a","slug":"mimic-iv-ext-cdm","version":"1.0","core_doi":"10.13026/6jkj-rb45","version_doi":"10.13026/2pfq-5b68","is_latest_version":false,"short_description":"A curated set of ED clinical decision making cases for four abdominal pathologies. Each case contains the exams required to diagnose including HPI, physical examination, laboratory tests, and imaging. Relevant treatment information is also included.","publish_date":"2024-05-17","title":"MIMIC-IV-Ext Clinical Decision Making: A MIMIC-IV Derived Dataset for Evaluation of Large Language Models on the Task of Clinical Decision Making for Abdominal Pathologies","abstract":"Clinical decision making is one of the most impactful parts of a physician's responsibilities and stands to benefit greatly from AI solutions such as large language models (LLMs). However, while many datasets exist to test the performance of AI models on constructed case vignettes, such as medical licensing exams, these tests fail to assess many skills that are necessary for deployment in a realistic clinical decision making environment. To understand how useful LLMs are in real-world settings, we must evaluate them in the wild, i.e. on real-world data under realistic conditions. To address this need, we have created a curated dataset based on the MIMIC-IV database, spanning 2400 real patient cases and four common abdominal pathologies: appendicitis, cholecystitis, diverticulitis, and pancreatitis. Each patient case contains the filtered and curated information necessary to arrive at the delivered diagnosis of the physician and can be used in an interactive manner to test the information gathering, synthesizing, and diagnostic capabilities of AI models.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":18024644,"compressed_storage_size":3809614,"resource_type":"Database","access_policy":"Credentialed","topics":["clinical decision making","emergency room diagnosis","abdominal pathologies","large language models","treatment plan"],"source_url":"https://physionet.org/content/mimic-iv-ext-cdm/1.0/"},{"public_project_uuid":"66e67e2d-e11d-4e3b-a5d7-716a61dcd5fe","slug":"temporal-respiratory-support","version":"1.0.0","core_doi":"10.13026/4nm5-h710","version_doi":"10.13026/0d8j-2w14","is_latest_version":false,"short_description":"A benchmark dataset offering hourly records over a 90-day period for 50,920 ICU subjects, including dynamic pulmonary function data and a spectrum of covariates for respiratory intervention analyses.","publish_date":"2024-05-31","title":"A Temporal Dataset for Respiratory Support in Critically Ill Patients","abstract":"We present a temporal benchmark dataset for clinical respiratory intervention tasks in intensive care unit (ICU) patients, derived from the MIMIC v2.2 dataset. The data consists of 50,920 adult ICU patients and includes 90-day hourly ventilation data, laboratory results, vital signs, and details of treatment interventions such as respiratory support interventions, vasopressor administration, and continuous renal replacement therapy. Additionally, it encompasses static variables such as ICU length of stay, discharge outcome, and death outcome. While there are existing open ICU datasets, the limited availability of benchmark datasets designed for clinical treatments in respiratory support motivated the development of the data. The data facilitates various tasks including time-series analysis, survival analysis, sequential decision-making tasks, and reinforcement learning.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":25713451615,"compressed_storage_size":452545922,"resource_type":"Database","access_policy":"Credentialed","topics":["oberservational data","time-series"],"source_url":"https://physionet.org/content/temporal-respiratory-support/1.0.0/"},{"public_project_uuid":"0966b491-f1fb-40f5-a96e-094157a72a34","slug":"me-llama","version":"1.0.0","core_doi":"10.13026/5158-hm11","version_doi":"10.13026/wwfd-2t39","is_latest_version":true,"short_description":"Me-LLaMA is a family of large language models for medical applications trained using clinical text with LLaMA2 models as the base. We release model weights for the foundation models as well as the chat-enhanced models.","publish_date":"2024-06-05","title":"Me-LLaMA: Foundation Large Language Models for Medical Applications","abstract":"Recent advancements in large language models (LLMs) such as ChatGPT and LLaMA have hinted at their potential to revolutionize medical applications, yet their application in clinical settings often reveals limitations due to a lack of specialized training on medical-specific data. In response to this challenge, this study introduces Me-LLaMA, a medical LLM family that includes foundation models – Me-LLaMA 13/70B, along with their chat-enhanced versions – Me-LLaMA 13/70B-chat, developed through continual pre-training and instruction tuning of LLaMA2 using large medical datasets. Our methodology leverages a comprehensive domain-specific data suite, including a large-scale, continual pre-training dataset with 129B tokens, an instruction tuning dataset with 214k samples, and we proposed a new medical evaluation benchmark (MIBE) across six critical medical tasks with 12 datasets. Our extensive evaluation using the MIBE shows that Me-LLaMA models achieve overall better performance than existing open-source medical LLMs in zero-shot, few-shot and supervised learning abilities. With task-specific instruction tuning, Me-LLaMA models outperform ChatGPT on 7 out of 8 datasets and GPT-4 on 5 out of 8 datasets. In addition, we investigated the catastrophic forgetting problem, and our results show that Me-LLaMA models outperform other open-source medical LLMs in mitigating this issue. Me-LLaMA is one of the largest open-source medical foundation LLMs that use both biomedical and clinical data. It exhibits superior performance across both general and medical tasks compared to other open-source medical LLMs, rendering it an attractive choice for medical AI applications.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":327973914243,"compressed_storage_size":280347856420,"resource_type":"Model","access_policy":"Credentialed","topics":["large language models"],"source_url":"https://physionet.org/content/me-llama/1.0.0/"},{"public_project_uuid":"49caa610-e530-435e-80c3-f98136eb2bae","slug":"grabmyo","version":"1.1.0","core_doi":"10.13026/cadk-kw71","version_doi":"10.13026/89dm-f662","is_latest_version":true,"short_description":"Open-access dataset of electromyogram (EMG) recordings collected from the wrist and forearm muscles of 43 people while they performed hand gestures.","publish_date":"2024-06-07","title":"Gesture Recognition and Biometrics ElectroMyogram (GRABMyo)","abstract":"We present the Gesture Recognition and Biometrics ElectroMyogram (GRABMyo) dataset, an open-access dataset of electromyogram (EMG) recordings collected from the wrist and forearm muscles while performing hand gestures. Data were collected from 43 healthy participants (age range: 24-35 years) on three different days while performing 16 hand and finger gestures in identical experimental sessions on each day. The GRABMyo dataset can be used for research on: 1) EMG-based biometrics for personal identification and verification, and 2) EMG-based gesture recognition for neurorehabilitation and home applications. The large sample size will provide sufficient power for establishing results, specifically for applications such as biometrics. Further, it will be useful for subject-independent applications such as generalized classification models for gesture recognition. The multiday recordings will provide results that will be consistent over a longer duration, crucial for the reliability of EMG-based wearable devices.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":10115481721,"compressed_storage_size":9820448666,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/grabmyo/1.1.0/"},{"public_project_uuid":"b82296bd-00c1-40f1-805f-25b0aa3a4276","slug":"rad-eval-x","version":"1.0.0","core_doi":"10.13026/72dp-f846","version_doi":"10.13026/tp88-q278","is_latest_version":true,"short_description":"The RadEvalX is a publicly available dataset developed similarly to the ReXVal dataset. RedEvalX focuses on radiologist evaluations of errors found in automatically generated radiology reports.","publish_date":"2024-06-18","title":"Radiology Report Generation Models Evaluation Dataset For Chest X-rays (RadEvalX)","abstract":"The Radiology Report Generation Models Evaluation Dataset For Chest X-rays (RadEvalX) is publicly available and developed similarly to the ReXVal dataset. Just like ReXVal, RadEvalX focuses on radiologist evaluations of errors found in automatically generated radiology reports. The dataset includes annotations from two board-certified radiologists, who identified clinically significant and clinically insignificant errors across eight different categories of errors. Compared to the ground-truth reports from the IU-Xray dataset, the evaluations were done on candidate radiology reports. For every 100 studies and corresponding ground-truth reports, the dataset contains one report generated using the M2Tr model from the corresponding X-ray image. The radiologists then annotated these reports. The primary purpose of this dataset is to assess the correlation between automated metrics and human radiologists' scoring, explore the limitations of automated metrics, and develop a model-based automated metric. This dataset has been created to support further research in medical artificial intelligence (AI), particularly in the field of radiology.
","license":{"name":"Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":147328,"compressed_storage_size":43231,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/rad-eval-x/1.0.0/"},{"public_project_uuid":"16379a15-401f-4bb0-93f4-e2a0d1e202b0","slug":"ehr-notes-qa-llms","version":"1.0.1","core_doi":"10.13026/mjgm-x360","version_doi":"10.13026/acga-ht95","is_latest_version":true,"short_description":"An LLM Benchmark for Real-World Clinical Practice Using Discharge Summaries","publish_date":"2024-06-26","title":"EHRNoteQA: An LLM Benchmark for Real-World Clinical Practice Using Discharge Summaries","abstract":"Discharge summaries in Electronic Health Records (EHRs) are crucial for clinical decision-making, but their length and complexity make information extraction challenging, especially when dealing with accumulated summaries across multiple patient admissions. Large Language Models (LLMs) show promise in addressing this challenge by efficiently analyzing vast and complex data. Existing benchmarks, however, fall short in properly evaluating LLMs' capabilities in this context, as they typically focus on single-note information or limited topics, failing to reflect the real-world inquiries required by clinicians. To bridge this gap, we introduce EHRNoteQA, a novel benchmark built on the MIMIC-IV EHR, comprising 962 different QA pairs each linked to distinct patients' discharge summaries. Every QA pair is initially generated using GPT-4 and then manually reviewed and refined by three clinicians to ensure clinical relevance. EHRNoteQA includes questions that require information across multiple discharge summaries and covers eight diverse topics, mirroring the complexity and diversity of real clinical inquiries.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":806274,"compressed_storage_size":191848,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/ehr-notes-qa-llms/1.0.1/"},{"public_project_uuid":"7abb67aa-7685-4e29-afc8-4d554f9550e2","slug":"mbrset","version":"1.0","core_doi":"10.13026/m10j-dj56","version_doi":"10.13026/qxpd-1y65","is_latest_version":true,"short_description":"mBRSET - a Mobile Brazilian Retinal Dataset","publish_date":"2024-06-26","title":"mBRSET, a Mobile Brazilian Retinal Dataset","abstract":"The Mobile Brazilian Retinal Dataset (mBRSET) presents a pioneering collection of retinal images captured via portable cameras, encompassing a diverse range of ethnic backgrounds from Itabuna, Bahia, Brazil. Comprising 5164 images from 1291 patients diagnosed with diabetes, this dataset is augmented with clinical and demographic metadata. Its significance lies in its provision as a resource for the development and validation of algorithms tailored for portable retinal cameras, which are increasingly being deployed, particularly in low and middle-income countries. By providing a dataset of retinal fundus photos captured via portable cameras, mBRSET offers an opportunity to develop and validate computer vision models for diabetic retinopathy. This resource has the potential to contribute to advancements in medical imaging and diagnostic technologies.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":2773006654,"compressed_storage_size":2580703408,"resource_type":"Database","access_policy":"Credentialed","topics":["retina","ophthalmology"],"source_url":"https://physionet.org/content/mbrset/1.0/"},{"public_project_uuid":"af208ffe-42bd-4f9a-9854-bb6438c00172","slug":"orchid","version":"2.0.0","core_doi":"10.13026/b1c0-3506","version_doi":"10.13026/w10c-6g08","is_latest_version":false,"short_description":"Multi-center dataset on organ procurement in the United States","publish_date":"2024-06-27","title":"Organ Retrieval and Collection of Health Information for Donation (ORCHID)","abstract":"There are well-documented inefficiencies and inequities in the current system of deceased donor organ transplantation. While much prior research has focused on designing better allocation systems to distribute donated organs, more can be done to study organ procurement, the process by which organs are recovered from deceased donors. Improvements to organ procurement can greatly increase the supply of a scarce public resource, improving health outcomes for the more than 100,000 patients currently waiting for an organ transplant. Such changes will also impact health equity, as a disproportionate number of waitlisted patients come from marginalized communities. So far, research into organ procurement has been hamstrung by a lack of transparent data. To address this pressing issue, we have collected the first ever multi-center dataset on organ procurement, covering ten years of clinical, financial, and administrative information from six organ procurement organizations (OPOs). This novel dataset can be used to identify opportunities to improve current practice, build tools to help OPOs make better decisions, and recommend policy changes to increase procurement.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":706790894,"compressed_storage_size":109629848,"resource_type":"Database","access_policy":"Restricted","topics":["organ procurement organizations","organ transplantation"],"source_url":"https://physionet.org/content/orchid/2.0.0/"},{"public_project_uuid":"f66de553-5139-44f9-9e57-374a77284245","slug":"mimic-iv-ext-cdm","version":"1.1","core_doi":"10.13026/6jkj-rb45","version_doi":"10.13026/ztyg-ah64","is_latest_version":true,"short_description":"A curated set of ED clinical decision making cases for four abdominal pathologies. Each case contains the exams required to diagnose including HPI, physical examination, laboratory tests, and imaging. Relevant treatment information is also included.","publish_date":"2024-07-08","title":"MIMIC-IV-Ext Clinical Decision Making: A MIMIC-IV Derived Dataset for Evaluation of Large Language Models on the Task of Clinical Decision Making for Abdominal Pathologies","abstract":"Clinical decision making is one of the most impactful parts of a physician's responsibilities and stands to benefit greatly from AI solutions such as large language models (LLMs). However, while many datasets exist to test the performance of AI models on constructed case vignettes, such as medical licensing exams, these tests fail to assess many skills that are necessary for deployment in a realistic clinical decision making environment. To understand how useful LLMs are in real-world settings, we must evaluate them in the wild, i.e. on real-world data under realistic conditions. To address this need, we have created a curated dataset based on the MIMIC-IV database, spanning 2400 real patient cases and four common abdominal pathologies: appendicitis, cholecystitis, diverticulitis, and pancreatitis. Each patient case contains the filtered and curated information necessary to arrive at the delivered diagnosis of the physician and can be used in an interactive manner to test the information gathering, synthesizing, and diagnostic capabilities of AI models.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":18048803,"compressed_storage_size":3820629,"resource_type":"Database","access_policy":"Credentialed","topics":["clinical decision making","abdominal pathologies","large language models","treatment plan","emergency room diagnosis"],"source_url":"https://physionet.org/content/mimic-iv-ext-cdm/1.1/"},{"public_project_uuid":"ec854b41-7a48-4bae-a40a-4cb4369cc9d8","slug":"radialog-instruct-dataset","version":"1.1.0","core_doi":"10.13026/pxb3-eh70","version_doi":"10.13026/40d2-wj86","is_latest_version":true,"short_description":"Image-based instruct data for Chest X-Ray understanding and analysis.","publish_date":"2024-07-12","title":"RaDialog Instruct Dataset","abstract":"Conversational AI tools that can generate and discuss clinically correct radiology reports for a given medical image have the potential to transform radiology. Such a human-in-the-loop radiology assistant could facilitate a collaborative diagnostic process, thus saving time and improving the quality of reports. Towards this goal, we introduce RaDialog, the first thoroughly evaluated and publicly available large vision-language model for radiology report generation and interactive dialog. To keep the conversational abilities of the underlying LLM, we propose a comprehensive, semi-automatically labeled, image-grounded instruct dataset for chest X-ray radiology tasks. The dataset includes a variety of tasks, such as report correction, summarization or finding prediction. By training with this dataset, our method achieves state-of-the-art clinical correctness in report generation and shows impressive abilities in interactive tasks such as correcting reports and answering questions, serving as a foundational step toward clinical dialog systems.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":885403298,"compressed_storage_size":133412049,"resource_type":"Database","access_policy":"Credentialed","topics":["medical image understaning","radiology chatbot","radiology report generation","radiology assistant","large vision-language models"],"source_url":"https://physionet.org/content/radialog-instruct-dataset/1.1.0/"},{"public_project_uuid":"cbeaab84-5a31-49c9-a601-4c3b2361d527","slug":"mimic-ext-mimic-cxr-vqa","version":"1.0.0","core_doi":"10.13026/my7r-v463","version_doi":"10.13026/deqx-d943","is_latest_version":true,"short_description":"We introduce MIMIC-Ext-MIMIC-CXR-VQA, a complex, diverse, and large-scale dataset designed for Visual Question Answering (VQA) tasks within the medical domain, focusing primarily on chest radiographs.","publish_date":"2024-07-19","title":"MIMIC-Ext-MIMIC-CXR-VQA: A Complex, Diverse, And Large-Scale Visual Question Answering Dataset for Chest X-ray Images","abstract":"We introduce MIMIC-Ext-MIMIC-CXR-VQA (i.e., Extended from MIMIC database), a complex, diverse, and large-scale dataset designed for Visual Question Answering (VQA) tasks within the medical domain, focusing primarily on chest radiographs. This dataset includes approximately 377K entries derived from the MIMIC-CXR-JPG, MIMIC-IV, and Chest ImaGenome datasets, all sourced from Physionet. It features questions generated from 48 unique templates across seven content types: presence, anatomy, attribute, abnormality, size, plane, and gender. Each template, developed under the guidance of a board-certified medical expert to ensure clinical relevance, addresses both standard content from previous medical VQA tasks and more complex scenarios involving set and logical operations. To further enhance linguistic diversity while maintaining a medical context, we implemented a paraphrasing strategy with an average of 16.5 paraphrases per template, developed through carefully designed prompts based on GPT-4.
\n\nThe primary aim of MIMIC-Ext-MIMIC-CXR-VQA is to serve as a comprehensive benchmark for evaluating medical VQA methodologies. However, the significance of this dataset extends far beyond just medical VQA benchmarking. It not only provides a foundational tool for developing and testing VQA methods but also acts as a valuable resource for instruction tuning of medical Vision-and-Language Models (VLMs), addressing the scarcity of medical instruction datasets. Furthermore, the integration of structured EHRs (i.e., MIMIC-IV) with our dataset, MIMIC-Ext-MIMIC-CXR-VQA, opens new avenues for the development of multi-modal AI frameworks that leverage both imaging and tabular modalities of patient records. By making this dataset publicly accessible, we aim to improve the understanding of medical images and stimulate further innovation within the realm of medical AI.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":350964129,"compressed_storage_size":28362291,"resource_type":"Database","access_policy":"Credentialed","topics":["multimodal","question answering","radiology","machine learning","deep learning","electronic health records","evaluation","benchmark","chest x-ray","visual question answering"],"source_url":"https://physionet.org/content/mimic-ext-mimic-cxr-vqa/1.0.0/"},{"public_project_uuid":"b96236ac-8946-441e-97ab-96c8f2021d89","slug":"mimic-cxr","version":"2.1.0","core_doi":"10.13026/s5dg-6s42","version_doi":"10.13026/4jqj-jw95","is_latest_version":true,"short_description":"Chest radiographs in DICOM format with associated free-text reports.","publish_date":"2024-07-23","title":"MIMIC-CXR Database","abstract":"The MIMIC Chest X-ray (MIMIC-CXR) Database v2.0.0 is a large publicly available dataset of chest radiographs in DICOM format with free-text radiology reports. The dataset contains 377,110 images corresponding to 227,835 radiographic studies performed at the Beth Israel Deaconess Medical Center in Boston, MA. The dataset is de-identified to satisfy the US Health Insurance Portability and Accountability Act of 1996 (HIPAA) Safe Harbor requirements. Protected health information (PHI) has been removed. The dataset is intended to support a wide body of research in medicine including image understanding, natural language processing, and decision support.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":5129964521754,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":["machine learning","mimic","computer vision","chest x-rays","radiology","natural language processing"],"source_url":"https://physionet.org/content/mimic-cxr/2.1.0/"},{"public_project_uuid":"4af1d6f8-ac77-493c-9506-d7f5a9f5ea38","slug":"mimiciv","version":"3.0","core_doi":"10.13026/07hj-2a80","version_doi":"10.13026/hxp0-hg59","is_latest_version":false,"short_description":"Large database of de-identified health information from patients admitted to Beth Israel Deaconess Medical Center","publish_date":"2024-07-23","title":"MIMIC-IV","abstract":"Retrospectively collected medical data has the opportunity to improve patient care through knowledge discovery and algorithm development. Broad reuse of medical data is desirable for the greatest public good, but data sharing must be done in a manner which protects patient privacy. Here we present Medical Information Mart for Intensive Care (MIMIC)-IV, a large deidentified dataset of patients admitted to the emergency department or an intensive care unit at the Beth Israel Deaconess Medical Center in Boston, MA. MIMIC-IV contains data for over 65,000 patients admitted to an ICU and over 200,000 patients admitted to the emergency department. MIMIC-IV incorporates contemporary data and adopts a modular approach to data organization, highlighting data provenance and facilitating both individual and combined use of disparate data sources. MIMIC-IV is intended to carry on the success of MIMIC-III and support a broad set of applications within healthcare.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":10659794737,"compressed_storage_size":10555320972,"resource_type":"Database","access_policy":"Credentialed","topics":["mimic","critical care","intensive care unit","machine learning"],"source_url":"https://physionet.org/content/mimiciv/3.0/"},{"public_project_uuid":"a24b5b5c-8280-4472-aa49-ff2bf1f829db","slug":"ehrxqa","version":"1.0.0","core_doi":"10.13026/0mst-es16","version_doi":"10.13026/bp5y-qt70","is_latest_version":true,"short_description":"We present EHRXQA, the first multi-modal EHR QA dataset combining structured patient records with aligned chest X-ray images. EHRXQA contains a comprehensive set of QA pairs covering image-related, table-related, and image+table-related questions.","publish_date":"2024-07-23","title":"EHRXQA: A Multi-Modal Question Answering Dataset for Electronic Health Records with Chest X-ray Images","abstract":"Electronic Health Records (EHRs), which contain patients' medical histories in various multi-modal formats, often overlook the potential for joint reasoning across imaging and table modalities underexplored in current EHR Question Answering (QA) systems. In this paper, we introduce EHRXQA, a novel multi-modal question answering dataset combining structured EHRs and chest X-ray images. To develop our dataset, we first construct two uni-modal resources: 1) The MIMIC-CXR-VQA dataset, our newly created medical visual question answering (VQA) benchmark, specifically designed to augment the imaging modality in EHR QA, and 2) EHRSQL (MIMIC-IV), a refashioned version of a previously established table-based EHR QA dataset. By integrating these two uni-modal resources, we successfully construct a multi-modal EHR QA dataset that necessitates both uni-modal and cross-modal reasoning. To address the unique challenges of multi-modal questions within EHRs, we propose a NeuralSQL-based strategy equipped with an external VQA API. This pioneering endeavor enhances engagement with multi-modal EHR sources and we believe that our dataset can catalyze advances in real-world medical scenarios such as clinical decision-making and research.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":230914310,"compressed_storage_size":26942942,"resource_type":"Database","access_policy":"Credentialed","topics":["machine learning","evaluation","visual question answering","question answering","electronic health records","benchmark","multi-modal question answering","deep learning","chest x-ray","ehr question answering","semantic parsing"],"source_url":"https://physionet.org/content/ehrxqa/1.0.0/"},{"public_project_uuid":"63982712-0cdd-492b-b6e9-f7fbc33c5d65","slug":"mimic-iv-ecg-ext-icd-labels","version":"1.0.0","core_doi":"10.13026/bd2x-mf68","version_doi":"10.13026/hdyc-1h77","is_latest_version":false,"short_description":"Dataset that links ECG records from MIMIC-IV-ECG to ED discharge and hospital discharge diagnoses, which enables to train general ECG prediction models based on clinical labels and facilitates the retrieval of further clinical metadata from MIMIC-IV.","publish_date":"2024-07-30","title":"MIMIC-IV-ECG-Ext-ICD: Diagnostic labels for MIMIC-IV-ECG","abstract":"The number of publicly available ECG datasets has increased tremendously in the past few years and several of these datasets have developed into widely used benchmarking datasets. However, most of them exhibit a common limitation, namely the reliance on retrospective annotation and a lack of clinical ground truth. This represents a serious limitation compared to closed in-hospital datasets. To circumvent this issue, we propose the MIMIC-IV-ECG-ICD dataset by linking the samples from the MIMIC-IV-ECG dataset to clinical ground truth from the MIMIC-IV dataset, in the form of ED and hospital discharge diagnoses. We release this derived dataset to foster further research on ECG-based prediction models with clinical ground truth and build a resource for benchmarking clinical ECG prediction models.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":344404865,"compressed_storage_size":39920080,"resource_type":"Database","access_policy":"Credentialed","topics":["electrocardiography","machine learning","mimic"],"source_url":"https://physionet.org/content/mimic-iv-ecg-ext-icd-labels/1.0.0/"},{"public_project_uuid":"34d4beb8-ed29-42a8-ad37-40491b407ab2","slug":"openox-repo","version":"1.0.1","core_doi":"10.13026/e821-1858","version_doi":"10.13026/2g7z-t345","is_latest_version":false,"short_description":"A repository of matched arterial oxygen and pulse oximeter readings obtained under controlled conditions, with high-frequency physiologic waveforms and skin color measurements.","publish_date":"2024-07-30","title":"OpenOximetry Repository","abstract":"The OpenOximetry Repository is a structured database designed to store clinical and laboratory pulse oximetry data and allows for consolidation of data sets held by collaborating organizations. Matched or independent readings of oxygen saturations, arterial blood gas samples, high frequency processed and unprocessed waveforms, and other physiologic metadata are supported (e.g. skin color readings). Contribution of data to the repository is encouraged; a data dictionary and standardized data collection protocols are provided to ensure consistent data archival and interpretation. The goal of this dataset is to enable researchers to interrogate the effects of physiologic and device variables on pulse oximeter performance.
\n\nA set of controlled desaturation studies is included in the initial release with ongoing updates planned as more data are available from both controlled lab desaturation studies and real world prospective clinical trials. Data captured during these studies may include paired pulse oximeter and arterial blood gas readings throughout a range of oxygen saturations, patient demographics and comorbidities, longitudinal quantitative and qualitative skin color measurement (using multiple methods at multiple anatomic locations), finger diameter, continuous ECG, arterial blood pressure, EtCO2/EtO2, unprocessed photoplethysmography (PPG), pH, PaCO2, hemoglobin, met-Hgb, CO-Hgb, respiratory rate and multiple other variables depending on the data source
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":2533046892,"compressed_storage_size":1584371188,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/openox-repo/1.0.1/"},{"public_project_uuid":"fc17ab3b-1afd-41b6-8d15-c10211c55511","slug":"radgraph2-radiology-reports","version":"1.0.0","core_doi":"10.13026/qptf-zh55","version_doi":"10.13026/q65y-9688","is_latest_version":true,"short_description":"RadGraph2 is a dataset of 800 chest radiology reports annotated using a fine-grained entity-relationship schema, which captures key findings as well as mentions of changes that occurred in comparison with the previous radiology studies.","publish_date":"2024-08-08","title":"RadGraph2: Tracking Findings Over Time in Radiology Reports","abstract":"RadGraph2 is a dataset of 800 chest radiology reports annotated using a fine-grained entity-relationship schema, which is an expanded version of the previously introduced RadGraph dataset. In contrast with the previous approaches and the original RadGraph, the new version of the used information extraction schema is designed to capture not only the key findings and their context but also the mentions of changes that occurred between the prior radiology examinations and the more recent study. These changes may include the appearance of new conditions affecting the patient, their progression, or the differences in the setup of the observed supporting devices. The information extracted from each report is represented in the form of a knowledge graph composed of clinically relevant entities and relations, which makes it easily amenable to automated processing. In addition to the dataset of manually labeled reports, we release more than 220,000 reports automatically annotated by our benchmark model. This model achieved an F1 micro performance of 0.88 and 0.74 on two differently sourced withheld test sets (from MIMIC-CXR-JPG and CheXpert, respectively). We believe that RadGraph2 could facilitate the development of clinically useful systems for the automated processing of radiology reports, particularly those reasoning about the evolution of a patient’s state over time.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":893588285,"compressed_storage_size":111158751,"resource_type":"Database","access_policy":"Credentialed","topics":["chest x-rays","radiology reports","disease progression","relation extraction","named entity recognition","information extraction"],"source_url":"https://physionet.org/content/radgraph2-radiology-reports/1.0.0/"},{"public_project_uuid":"f0c20c62-9fc0-40d1-8728-cbbf4a26faf2","slug":"inspire","version":"1.3","core_doi":"10.13026/dv8j-5412","version_doi":"10.13026/46m4-f655","is_latest_version":false,"short_description":"A public dataset that contains information related to surgery, anesthesia, laboratory results, medications, diagnosis, and outcomes from 50% of the patients who received surgery at Seoul National University Hospital between 2011 and 2020.","publish_date":"2024-08-12","title":"INSPIRE, a publicly available research dataset for perioperative medicine","abstract":"We present the INSPIRE dataset, a publicly available research dataset in perioperative medicine, which includes approximately 130,000 cases (50% of all surgical cases) who underwent anesthesia for surgery at an academic institution in South Korea between 2011 and 2020. This comprehensive dataset includes patient characteristics such as age, sex, American Society of Anesthesiologists physical status classification, diagnosis, surgical procedure code, department, and type of anesthesia. It also includes vital signs in the operating theatre, general wards, and intensive care units (ICUs), laboratory results from six months before admission to six months after discharge, and medication during hospitalization. Complications include total hospital and ICU length of stay and in-hospital death. We hope this dataset will inspire collaborative research and development in perioperative medicine and serve as a reproducible external validation dataset to improve surgical outcomes.
","license":{"name":"Korea Credentialed Health Data License-1-0-0"},"dua":{"name":"Korea Credentialed Health Data Agreement-1-0-0"},"main_storage_size":4778369069,"compressed_storage_size":540276524,"resource_type":"Database","access_policy":"Credentialed","topics":["surgery","open dataset","multi-center","perioperative medicine"],"source_url":"https://physionet.org/content/inspire/1.3/"},{"public_project_uuid":"b070f5ae-a36c-4a88-8379-10ac6aa44aea","slug":"shdb-af","version":"1.0.0","core_doi":"10.13026/f5mp-gx65","version_doi":"10.13026/10mk-y852","is_latest_version":false,"short_description":"Holter ECG database from Japan, containing data from 100 unique patients with paroxysmal AF including expert annotations of Supraventricular arrhythmias at the beat level.","publish_date":"2024-08-12","title":"SHDB-AF: a Japanese Holter ECG database of atrial fibrillation","abstract":"Saitama Heart Database Atrial Fibrillation (SHDB-AF) is a novel open-sourced Holter ECG database from Japan, containing data from 100 unique patients with paroxysmal atrial fibrillation. The dataset contains raw ECG recordings with manually annotated rhythm at the beat level performed by a fellow in cardiology. The dataset was collected as part of a study evaluating the generalization performance of a deep learning atrial fibrillation event detection model across different distribution shifts.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":6935572214,"compressed_storage_size":6089804447,"resource_type":"Database","access_policy":"Open","topics":["ecg","atrial fibrillation","holters"],"source_url":"https://physionet.org/content/shdb-af/1.0.0/"},{"public_project_uuid":"a57c0d6c-bfb0-494a-8da8-cab006914039","slug":"brazilian-ophthalmological","version":"1.0.1","core_doi":"10.13026/z9zv-g542","version_doi":"10.13026/1pht-2b69","is_latest_version":false,"short_description":"This is the first Brazilian Multilabel Ophthalmological Dataset with demographic information and retinal photos labeled images according to anatomical parameters, quality control, and presumed diagnosis.","publish_date":"2024-08-14","title":"A Brazilian Multilabel Ophthalmological Dataset (BRSET)","abstract":"The Brazilian Multilabel Ophthalmological Dataset (BRSET) is a multi-labeled ophthalmological dataset designed to improve scientific community development and validate machine learning models. In ophthalmology, ancillary exams support medical decisions and can be used to develop algorithms; however, the availability and representativeness of ophthalmological datasets are limited. This dataset consists of 16,266 images from 8,524 Brazilian patients. Demographics, macula, optic disc, and vessels anatomical parameters, focus, illumination, image field, and artifacts as quality control, and multi-labels are included alongside color fundus retinal photos. This dataset enables computer vision models to predict demographic characteristics and multi-label disease classification using retinal fundus photos.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":16711737716,"compressed_storage_size":13072442556,"resource_type":"Database","access_policy":"Credentialed","topics":["dataset","retina","ophthalmology"],"source_url":"https://physionet.org/content/brazilian-ophthalmological/1.0.1/"},{"public_project_uuid":"b23a51bd-c1d3-4b9a-b094-89450395c3b0","slug":"rexpref-prior","version":"1.0.0","core_doi":"10.13026/hwzp-zq96","version_doi":"10.13026/t13x-4r94","is_latest_version":true,"short_description":"We propose ReXPref-Prior, an adapted version of MIMIC-CXR where GPT-4 has removed references to prior exams from both findings and impression sections of chest X-ray reports.","publish_date":"2024-08-14","title":"ReXPref-Prior: A MIMIC-CXR Preference Dataset for Reducing Hallucinated Prior Exams in Radiology Report Generation","abstract":"Generative vision-language models have exciting potential implications for radiology report generation, but unfortunately such models are also known to produce hallucinations and other nonsensical statements. For example, radiology report generation models regularly hallucinate prior exams, making statements such as “The lungs are hyperinflated with emphysematous changes as seen on prior CT” despite not having access to any prior exam. To address this shortcoming, we propose ReXPref-Prior, an adapted version of MIMIC-CXR where GPT-4 has removed references to prior exams from both findings and impression sections of chest X-ray reports. We expect ReXPref-Prior will be useful for training models that hallucinate prior exams less frequently, through techniques such as direct preference optimization. Additionally, ReXPref-Prior’s validation and test sets can be used as a new benchmark for evaluating report generation models.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":21038899,"compressed_storage_size":3634974,"resource_type":"Database","access_policy":"Credentialed","topics":["hallucination","reinforcement learning","chest x-rays"],"source_url":"https://physionet.org/content/rexpref-prior/1.0.0/"},{"public_project_uuid":"145b49f3-bed2-4496-a9e1-eb1e68988073","slug":"ehrcon-consistency-of-notes","version":"1.0.0","core_doi":"10.13026/x1ea-np24","version_doi":"10.13026/2nb5-nf74","is_latest_version":false,"short_description":"Dataset for Checking Consistency between Unstructured Notes and Structured Tables in Electronic Health Records","publish_date":"2024-08-19","title":"EHRCon: Dataset for Checking Consistency between Unstructured Notes and Structured Tables in Electronic Health Records","abstract":"Electronic Health Records (EHRs) are integral for storing comprehensive patient medical records, combining structured data (e.g., medications) with detailed clinical notes (e.g., physician notes). These elements are essential for straightforward data retrieval and provide deep, contextual insights into patient care. However, they often suffer from discrepancies due to unintuitive EHR system designs and human errors, posing serious risks to patient safety. To address this, we developed EHRCon, a new dataset and task specifically designed to ensure data consistency between structured tables and unstructured notes in EHRs. EHRCon was crafted in collaboration with healthcare professionals using the MIMIC-III EHR dataset, and includes manual annotations of 3,943 entities across 105 clinical notes checked against database entries for consistency. EHRCon has two versions, one using the original MIMIC-III schema, and another using the OMOP CDM schema, in order to increase its applicability and generalizability.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":8269371,"compressed_storage_size":1526366,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/ehrcon-consistency-of-notes/1.0.0/"},{"public_project_uuid":"de24126a-33bd-4e79-bc8c-ce0db2815aee","slug":"encode-skin-color","version":"1.0.0","core_doi":"10.13026/33g5-v628","version_doi":"10.13026/mcgk-1s42","is_latest_version":true,"short_description":"A prospective collected EHR-linked skin tone measurements database in OMOP format with emphasis on pulse oximetry disparities.","publish_date":"2024-08-22","title":"ENCoDE, mEasuring skiN Color to correct pulse Oximetry DisparitiEs: skin tone and clinical data from a prospective trial on acute care patients.","abstract":"The wide adoption of pulse oximeters has given clinicians an easy, non-invasive way to measure arterial oxygen saturation. However, evidence suggests that pulse oximeter measurements have a deeper discrepancy in patients of darker skin tones compared to their lighter counterparts. It has been hypothesized that skin tone is the root cause of this phenomenon. However, skin tone as a medical concept has not been extensively studied in acute care.
\n\nOur study included patients admitted to Duke University Hospital with pulse oximetry recorded up to 5 minutes prior to arterial blood gas (ABG) measurements. Skin tone was measured across sixteen body locations using administered visual scales (Fitzpatrick, Monk Skin Tone, and Von Luschan), reflectance colorimetry (Delfin SkinColorCatch), and reflectance spectrophotometry (Konica Minolta CM-700D, Variable Spectro 1). IPhone SE 2020 and Google Pixel 4 (Android) image data are available for non-biometric body locations.
\n\nOne hundred twenty-eight patients are enrolled in this study. A total of 167 skin tone variables and two temperature variables are collected per body location, excluding images, together with ten non-biometric body location images per patient and the associated electronic health record (EHR) data. The ENCoDE project is a comprehensive EHR-linked skin tone database to combat skin tone associate disparities.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1033291338,"compressed_storage_size":691280593,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/encode-skin-color/1.0.0/"},{"public_project_uuid":"1ec1a0c8-7b2b-441b-9f2a-6b0ba52dc4cc","slug":"butppg","version":"2.0.0","core_doi":"10.13026/gxwp-p940","version_doi":"10.13026/tn53-8153","is_latest_version":true,"short_description":"BUT PPG is a database created for the purpose of evaluating PPG signal quality and estimation of heart rate. The data comprises 3,888 10s recordings of PPGs recorded by smartphone and associated ECG and ACC signals and annotations.","publish_date":"2024-08-23","title":"Brno University of Technology Smartphone PPG Database (BUT PPG)","abstract":"Brno University of Technology Smartphone PPG Database (BUT PPG) is a database created by the cardiology team at the Department of Biomedical Engineering, Brno University of Technology, for the purpose of evaluating PPG quality and estimation of heart rate (HR). The data comprises 3,888 10-second recordings of PPGs and associated ECG signals used for determination of reference HR. All of the signals contain QRS complex positions, most signals contain moreover accelerometric data (ACC), and single-value annotations of blood pressure, blood oxygen saturation and glycaemia. The data were collected from 50 subjects (25 female, 25 male) aged between 19 to 76 years at rest and during various types of movement. Recordings were carried out between August 2020 and December 2021. PPG data were collected by smartphones Xiaomi Mi9 and Huawei P20 Pro with sampling frequency of 30 Hz. Reference ECG signals and ACC data were recorded using a mobile recorder Bittium Faros 360 or 180 with a sampling frequency of 1,000 Hz (ECG) and 100 Hz (ACC). Each PPG signal includes annotation of quality and reference HR. PPG signal quality is indicated binary: 1 indicates good quality for HR estimation, 0 indicates signals where HR cannot be detected reliably and thus these signals are unsuitable for any analysis.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":213136194,"compressed_storage_size":90951417,"resource_type":"Database","access_policy":"Open","topics":["heart rate","acc","artificial intelligence","signal quality assessment","annotations","accelerometric data","ppg","electrocardiogram","photoplethysmography","ecg"],"source_url":"https://physionet.org/content/butppg/2.0.0/"},{"public_project_uuid":"162f2bc7-2412-4332-b8dc-d36ada81612b","slug":"multimodal-surgery-anesthesia","version":"1.0","core_doi":"10.13026/gzmm-5h49","version_doi":"10.13026/gs4v-4q80","is_latest_version":true,"short_description":"Multimodal physiological indices collected during surgery when patients were under anesthesia","publish_date":"2024-08-23","title":"Multimodal Physiological Indices During Surgery Under Anesthesia","abstract":"Monitoring nociception, the flow of information associated with harmful stimuli through the nervous system even during unconsciousness, is critical for proper anesthesia care during surgery. Currently, this is done by tracking heart rate and blood pressure by eye. Monitoring objectively a patient’s nociceptive state remains a challenge, causing drugs to often be over- or under-dosed intraoperatively. Inefficient management of surgical nociception may lead to more complex post-operative pain management and side effects such as post-operative cognitive dysfunction, particularly in elderly patients. We collected a comprehensive and multi-sensor prospective observational dataset focused on surgical nociception (101 surgeries, 18,582 minutes, 49,878 nociceptive stimuli), including annotations of all nociceptive stimuli occurring during surgery and medications administered. Using this dataset, we developed indices of autonomic nervous system activity based on physiologically and statistically rigorous point process representations of cardiac action potentials and sweat gland activity. Next, we constructed highly interpretable supervised and unsupervised models with appropriate inductive biases that quantify surgical nociception throughout surgery. Our models track nociceptive stimuli more accurately than existing nociception monitors.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":375369919,"compressed_storage_size":212219313,"resource_type":"Database","access_policy":"Restricted","topics":["nociception","anesthesia"],"source_url":"https://physionet.org/content/multimodal-surgery-anesthesia/1.0/"},{"public_project_uuid":"28630e69-c9c5-4f79-96de-ec2bd712c87d","slug":"cxr-cardiomegaly","version":"1.0.0","core_doi":"10.13026/4evw-jd69","version_doi":"10.13026/kfpv-zm25","is_latest_version":true,"short_description":"Automatically extracted cardiomegaly biomarkers - cardiothoracic ratio (CTR) and cardiopulmonary area ratio (CPAR) - for all posterior-anterior chest x-ray scans in MIMIC-CXR/MIMIC-CXR-JPG.","publish_date":"2024-08-23","title":"Image-derived cardiomegaly biomarker values for 96K chest X-rays in MIMIC-CXR/MIMIC-CXR-JPG","abstract":"Cardiomegaly is a condition characterized by an abnormal enlargement of the heart, its identification is of paramount importance as it associate with a wide range of cardiac conditions. It is primary identified via the cardiothoracic ratio (CTR), however this metric can be inaccurate as it is affect by external factors such as breathing and body position. Multimodal approaches could mitigate these limitations by integrating non-imaging data, however reliable and explainable integration of imaging and non-imaging data remains a significant challenge. While this database does not directly use multimodal data, it hopes to tackle this challenge by extracting cardiomegaly biomarkers (CTR and cardiopulmonary area ratio) from chest X-rays. Thus encapsulating the relevant imaging information into individual datapoints, allowing easy integration of ‘imaging’ data with non-imaging data for more reliable diagnostic tools. The values were extracted from over 93,000 posterior-anterior MIMIC-CXR scans using detection and segmentation neural networks, tuned for cardiac and pulmonary identification.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":12450834,"compressed_storage_size":6974865,"resource_type":"Database","access_policy":"Open","topics":["cpar","biomarkers","ctr","mimic-cxr","cardiomegaly"],"source_url":"https://physionet.org/content/cxr-cardiomegaly/1.0.0/"},{"public_project_uuid":"65569c6c-2bca-4e8e-9502-5bd7f66577f8","slug":"mimic-iv-ecg-ext-icd-labels","version":"1.0.1","core_doi":"10.13026/bd2x-mf68","version_doi":"10.13026/ypt5-9d58","is_latest_version":true,"short_description":"Dataset that links ECG records from MIMIC-IV-ECG to ED discharge and hospital discharge diagnoses, which enables to train general ECG prediction models based on clinical labels and facilitates the retrieval of further clinical metadata from MIMIC-IV.","publish_date":"2024-08-30","title":"MIMIC-IV-ECG-Ext-ICD: Diagnostic labels for MIMIC-IV-ECG","abstract":"The number of publicly available ECG datasets has increased tremendously in the past few years and several of these datasets have developed into widely used benchmarking datasets. However, most of them exhibit a common limitation, namely the reliance on retrospective annotation and a lack of clinical ground truth. This represents a serious limitation compared to closed in-hospital datasets. To circumvent this issue, we propose the MIMIC-IV-ECG-Ext-ICD dataset by linking the samples from the MIMIC-IV-ECG dataset to clinical ground truth from the MIMIC-IV dataset, in the form of ED and hospital discharge diagnoses. We release this derived dataset to foster further research on ECG-based prediction models with clinical ground truth and build a resource for benchmarking clinical ECG prediction models.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":338915072,"compressed_storage_size":37384702,"resource_type":"Database","access_policy":"Credentialed","topics":["mimic","machine learning","electrocardiography"],"source_url":"https://physionet.org/content/mimic-iv-ecg-ext-icd-labels/1.0.1/"},{"public_project_uuid":"6c62c66d-3a86-42b4-9be2-e5db14affaec","slug":"sicdb","version":"1.0.8","core_doi":"10.13026/ezs8-6v88","version_doi":"10.13026/8m72-6j83","is_latest_version":true,"short_description":"The SICdb dataset, version 1.0.8 contains 27350 admissions to an ICU in an Austrian tertiary care institution.","publish_date":"2024-09-10","title":"Salzburg Intensive Care database (SICdb), a freely accessible intensive care database","abstract":"The SICdb dataset offers insights into over 27 thousand intensive care admissions, including therapies and data on preceding surgeries. Data were collected between 2013 and 2021 from four different intensive care units at the University Hospital Salzburg, having more than 3 thousand intensive care admissions per year on 41 beds. The dataset is deidentified and contains, amongst others, case information, vital signs, laboratory results and medication data. SICdb provides both aggregated once-per-hour and highly granular once-per-minute data, making it suitable for computational and machine learning-based research.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":2379847584,"compressed_storage_size":2377018665,"resource_type":"Database","access_policy":"Contributor Review","topics":["clinical","machine learning","critical care","intensive care","open data"],"source_url":"https://physionet.org/content/sicdb/1.0.8/"},{"public_project_uuid":"1cef94d3-a7ad-4a35-9ece-e65e1b28fba2","slug":"labelled-notes-hospital-course","version":"1.0.0","core_doi":"10.13026/fh2q-4148","version_doi":"10.13026/vr29-4d92","is_latest_version":false,"short_description":"This dataset presents a collection of preprocessed and labeled clinical notes derived from \"MIMIC-IV-Note\", and aims to facilitate the development of ML models focused on summarizing brief hospital courses (BHC) from clinical notes.","publish_date":"2024-09-10","title":"MIMIC-IV-Ext-BHC: Labeled Clinical Notes Dataset for Hospital Course Summarization","abstract":"This dataset presents a curated collection of preprocessed and labeled clinical notes derived from the MIMIC-IV-Note database. The primary aim of this resource is to facilitate the development and training of machine learning models focused on summarizing brief hospital courses (BHC) from clinical discharge notes.
\n\nThe dataset contains 270,033 meticulously cleaned and standardized clinical notes containing an average token length of 2,267, ensuring usability for machine learning (ML) applications. Each clinical note is paired with a corresponding BHC summary, providing a robust foundation for supervised learning tasks. The preprocessing pipeline employed uses regular expressions to address common issues in the raw clinical text, such as special characters, extraneous whitespace, inconsistent formatting, and irrelevant text, to produce a high-quality, structured dataset with separated clinical note sections through appropriate headings.
\n\nBy offering this resource, we aim to support healthcare professionals and researchers in their efforts to enhance patient care through the automation of BHC summarization. This dataset is ideal for exploring various NLP techniques, developing predictive models, and improving the efficiency and accuracy of clinical documentation practices. We invite the research community to utilize this dataset to advance the field of medical informatics and contribute to better health outcomes.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":2724293236,"compressed_storage_size":862356297,"resource_type":"Database","access_policy":"Credentialed","topics":["clinical notes","machine learning","natural language processing","text summarization","brief hospital course"],"source_url":"https://physionet.org/content/labelled-notes-hospital-course/1.0.0/"},{"public_project_uuid":"6d7e8d60-6e32-456b-9f5e-78cd3362b566","slug":"music-sudden-cardiac-death","version":"1.0.0","core_doi":"10.13026/fa8p-he52","version_doi":"10.13026/cec2-9w70","is_latest_version":false,"short_description":"The MUSIC study is a prospective, multicentre, longitudinal study designed to assess risk predictors of cardiac mortality and sudden cardiac death in ambulatory patients with chronic heart failure.","publish_date":"2024-09-11","title":"MUSIC (Sudden Cardiac Death in Chronic Heart Failure)","abstract":"The MUSIC (MUerte Subita en Insuficiencia Cardiaca) study is a prospective, multicentre, longitudinal study designed to assess risk predictors of cardiac mortality and sudden cardiac death (SCD) in ambulatory patients with chronic heart failure (CHF).
\n\nThe study population consisted of 992 patients with CHF consecutively enrolled from the specialized HF clinics of eight University Spanish Hospitals between April 2003 and December 2004, and followed up for a median of 44 months (until November 2008). All patients had a 12-lead electrocardiogram (ECG), a 24 h, 2-(4%) or 3-lead (96%) Holter ECG, chest X-ray, echocardiography, and blood laboratory parameters performed at enrolment.
\n\nPrimary outcomes were cardiac death, either sudden cardiac death (SCD) or pump failure death (PFD) at the end of the follow-up period.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":96891031656,"compressed_storage_size":46665998758,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/music-sudden-cardiac-death/1.0.0/"},{"public_project_uuid":"fff7e608-ebd6-4906-b626-414d197a59d0","slug":"multimodal-emergency-benchmark","version":"1.0.0","core_doi":"10.13026/fm1f-2558","version_doi":"10.13026/p90d-vd84","is_latest_version":true,"short_description":"MIMIC-IV-ext-MDS-ED proposes a dataset to benchmark multimodal decision support in the emergency department. It features multimodal input (including ECG waveforms) and a comprehensive set of prediction targets (diagnoses and deterioration prediction)","publish_date":"2024-09-12","title":"MIMIC-IV-Ext-MDS-ED: Multimodal Decision Support in the Emergency Department - a Benchmark Dataset for Diagnoses and Deterioration Prediction in Emergency Medicine","abstract":"Measurable progress in the development of medical decision support systems has been hindered by a lack of comprehensive datasets. Many available datasets focus on narrow prediction tasks and do not include a diverse range of data types, which limits their effectiveness in real-world clinical settings. This issue is particularly critical in emergency care, where accurate and timely diagnoses and the ability to predict patient deterioration are essential.
\n\nTo address these challenges, we present a new dataset derived from MIMIC-IV, created specifically for benchmarking multimodal decision support systems in emergency departments. This dataset includes data from the first 1.5 hours after the patient's arrival, covering demographics, biometrics, vital signs (including trends), lab results (including trends), and ECG waveforms. It allows for the evaluation of predictive models across a broad spectrum of clinical conditions, including both cardiac and non-cardiac conditions (1428 ICD-10 codes), as well as clinical deterioration measures (15 labels covering 6 clinical deterioration conditions, ICU admission at two horizons, and mortality at 7 different time horizons).
\n\nThe integration of diverse data types aims to enhance the clinical relevance and robustness of decision support systems, facilitating more accurate and timely predictions in acute care scenarios. We release this dataset to encourage further research and innovation in emergency medicine and to provide a resource for the reliable benchmarking of multimodal AI models in the field.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":587734798,"compressed_storage_size":45538585,"resource_type":"Database","access_policy":"Credentialed","topics":["multimodal","emergency department","benchmark","diagnoses prediction","ecg","deterioration prediction"],"source_url":"https://physionet.org/content/multimodal-emergency-benchmark/1.0.0/"},{"public_project_uuid":"398a10ac-e10c-4f88-817d-e36d4f009363","slug":"dreamt","version":"1.0.1","core_doi":"10.13026/dztc-dv77","version_doi":"10.13026/nktq-0858","is_latest_version":false,"short_description":"Dataset for Real-time sleep stage EstimAtion using Multisensor wearable Technology","publish_date":"2024-09-13","title":"DREAMT: Dataset for Real-time sleep stage EstimAtion using Multisensor wearable Technology","abstract":"Sleep is an intrinsic part of human life, and recent advancements in wearable technology and machine learning have promised continuous and non-invasive methods of tracking sleep health and patterns, providing an important facet to a more holistic understanding of well-being. However, it is still challenging to achieve consistent and reliable real-time estimates of sleep stages using only smartwatches. This is especially true for individuals with irregular sleep patterns or sleep disorders. A major contributing factor is the distinct lack of publicly accessible, large-scale datasets that allow researchers and engineers to validate their wearable sleep staging algorithms against a population with diverse sleep patterns. Here, we present DREAMT, Dataset for Real-time sleep stage EstimAtion using Multisensor wearable Technology, a new dataset collected from 100 participants, which includes high-resolution signals from a smartwatch, expert sleep technician-annotated sleep stage labels, and clinical metadata related to sleep health and disorders.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":14108970053,"compressed_storage_size":1422765357,"resource_type":"Database","access_policy":"Restricted","topics":["biomedical","sleep disorders","time series classification","wearable"],"source_url":"https://physionet.org/content/dreamt/1.0.1/"},{"public_project_uuid":"5ff6259d-ab56-4091-96e8-da0c9d674ac5","slug":"cps-dataset-sleep","version":"1.0.0","core_doi":"10.13026/1949-4b29","version_doi":"10.13026/sxs0-h317","is_latest_version":true,"short_description":"This dataset includes polysomnographic sleep recordings from a study on sleep-related arousal diagnostics, featuring raw and derived data channels, annotated event types, and questionnaire data.","publish_date":"2024-09-18","title":"Comprehensive Polysomnography (CPS) Dataset: A Resource for Sleep-Related Arousal Research","abstract":"The CPS dataset, assembled at Klinikum Esslingen, Germany, encompasses 113 diagnostic polysomnographic sleep recordings. These recordings include up to 36 raw and 23 derived data channels, alongide 81 types of annotated events for each participant, supplemented by data from various questionnaires. The dataset was collected during 2021-2022 for the medical study "Computer-aided diagnostics of sleep-related arousals on the basis of pulse wave analyses" in a collaborative effort between Klinikum Esslingen, IT-Designers Gruppe and NRI Medizintechnik GmbH, all based in Germany. The dataset may be used for research that is in-line with the original study goals (see more in section "Usage Notes").
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":120995650585,"compressed_storage_size":111771533133,"resource_type":"Database","access_policy":"Credentialed","topics":["sleep disorders","polysomnography","machine learning in healthcare","sleep arousal diagnostics","pulse wave analysis"],"source_url":"https://physionet.org/content/cps-dataset-sleep/1.0.0/"},{"public_project_uuid":"557b4d78-70db-4205-8296-a36119c0917c","slug":"vtac","version":"1.0","core_doi":"10.13026/z4f3-1f07","version_doi":"10.13026/8td2-g363","is_latest_version":false,"short_description":"VTaC is an annotated ventricular tachycardia (VT) arrhythmia alarm database containing over 5,000 waveform recordings with VT alarms from ICU monitors, with each alarm labeled as either true or false by at least two human expert annotators.","publish_date":"2024-10-01","title":"VTaC: A Benchmark Dataset of Ventricular Tachycardia Alarms from ICU Monitors","abstract":"False arrhythmia alarms are a persistent problem in intensive care units despite considerable effort from industrial and academic algorithm developers. Among the various arrhythmias, ventricular tachycardia (VT) is particularly challenging to detect accurately, as achieving both high sensitivity and high positive predictivity has proven difficult. We present an annotated VT alarm database, VTaC (Ventricular Tachycardia annotated alarms from ICUs) consisting of over 5,000 waveform recordings with VT alarms triggered by bedside monitors in the ICUs. Each VT alarm in the dataset was labeled as true or false by at least two independent human expert annotators. The dataset comprises data collected from ICUs in three major US hospitals and includes data from three leading bedside monitor manufacturers, providing a diverse and representative collection of VT alarm waveform data. Each waveform recording comprises at least two electrocardiogram (ECG) leads and one or more pulsatile waveforms, such as photoplethysmogram (PPG or PLETH) and arterial blood pressure (ABP) waveforms.
","license":{"name":"Creative Commons Attribution-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":4331092450,"compressed_storage_size":2866704109,"resource_type":"Database","access_policy":"Open","topics":["machine learning","icu false alarms","benchmark dataset","ventricular tachycardia","arrhythmia"],"source_url":"https://physionet.org/content/vtac/1.0/"},{"public_project_uuid":"809172de-e495-41df-89ff-34a88ac216a3","slug":"labelled-notes-hospital-course","version":"1.1.0","core_doi":"10.13026/fh2q-4148","version_doi":"10.13026/41et-8342","is_latest_version":false,"short_description":"This dataset presents a collection of preprocessed and labeled clinical notes derived from \"MIMIC-IV-Note\", and aims to facilitate the development of ML models focused on summarizing brief hospital courses (BHC) from clinical notes.","publish_date":"2024-10-10","title":"MIMIC-IV-Ext-BHC: Labeled Clinical Notes Dataset for Hospital Course Summarization","abstract":"This dataset presents a curated collection of preprocessed and labeled clinical notes derived from the MIMIC-IV-Note database. The primary aim of this resource is to facilitate the development and training of machine learning models focused on summarizing brief hospital courses (BHC) from clinical discharge notes.
\n\nThe dataset contains 270,033 meticulously cleaned and standardized clinical notes containing an average token length of 2,267, ensuring usability for machine learning (ML) applications. Each clinical note is paired with a corresponding BHC summary, providing a robust foundation for supervised learning tasks. The preprocessing pipeline employed uses regular expressions to address common issues in the raw clinical text, such as special characters, extraneous whitespace, inconsistent formatting, and irrelevant text, to produce a high-quality, structured dataset with separated clinical note sections through appropriate headings.
\n\nBy offering this resource, we aim to support healthcare professionals and researchers in their efforts to enhance patient care through the automation of BHC summarization. This dataset is ideal for exploring various NLP techniques, developing predictive models, and improving the efficiency and accuracy of clinical documentation practices. We invite the research community to utilize this dataset to advance the field of medical informatics and contribute to better health outcomes.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":2726467690,"compressed_storage_size":863062772,"resource_type":"Database","access_policy":"Credentialed","topics":["machine learning","brief hospital course","clinical notes","natural language processing","text summarization"],"source_url":"https://physionet.org/content/labelled-notes-hospital-course/1.1.0/"},{"public_project_uuid":"08848569-9a6c-4333-a9c6-b89482ea0cc5","slug":"mimiciv","version":"3.1","core_doi":"10.13026/07hj-2a80","version_doi":"10.13026/kpb9-mt58","is_latest_version":true,"short_description":"Large database of de-identified health information from patients admitted to Beth Israel Deaconess Medical Center","publish_date":"2024-10-11","title":"MIMIC-IV","abstract":"Retrospectively collected medical data has the opportunity to improve patient care through knowledge discovery and algorithm development. Broad reuse of medical data is desirable for the greatest public good, but data sharing must be done in a manner which protects patient privacy. Here we present Medical Information Mart for Intensive Care (MIMIC)-IV, a large deidentified dataset of patients admitted to the emergency department or an intensive care unit at the Beth Israel Deaconess Medical Center in Boston, MA. MIMIC-IV contains data for over 65,000 patients admitted to an ICU and over 200,000 patients admitted to the emergency department. MIMIC-IV incorporates contemporary data and adopts a modular approach to data organization, highlighting data provenance and facilitating both individual and combined use of disparate data sources. MIMIC-IV is intended to carry on the success of MIMIC-III and support a broad set of applications within healthcare.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":10656227926,"compressed_storage_size":10551747784,"resource_type":"Database","access_policy":"Credentialed","topics":["critical care","intensive care unit","mimic","machine learning"],"source_url":"https://physionet.org/content/mimiciv/3.1/"},{"public_project_uuid":"3c7539af-6e29-440d-8f55-75c4e983d0a7","slug":"multimodal-dental-dataset","version":"1.1.0","core_doi":"10.13026/bxzm-dg78","version_doi":"10.13026/h1tt-fc69","is_latest_version":true,"short_description":"A new dental dataset that contains 169 patients, three commonly used dental image models, and images of various health conditions of the oral cavity.","publish_date":"2024-10-11","title":"A multimodal dental dataset facilitating machine learning research and clinic services","abstract":"Oral diseases affect nearly 3.5 billion people, with the majority residing in low- and middle-income countries. Due to limited healthcare resources, many individuals are unable to access proper oral healthcare services. Image-based machine learning technology is one of the most promising approaches to improving oral healthcare services and reducing patient costs. Openly accessible datasets play a crucial role in facilitating the development of machine learning techniques. However, existing dental datasets have limitations such as a scarcity of Cone Beam Computed Tomography (CBCT) data, lack of matched multi-modal data, and insufficient complexity and diversity of the data. This project addresses these challenges by providing a dataset that includes 329 CBCT images from 169 patients, multi-modal data with matching modalities, and images representing various oral health conditions.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":113744130974,"compressed_storage_size":50934191452,"resource_type":"Database","access_policy":"Contributor Review","topics":[],"source_url":"https://physionet.org/content/multimodal-dental-dataset/1.1.0/"},{"public_project_uuid":"10f309b7-fd9b-401c-b8ae-73ff631763d3","slug":"meddec","version":"1.0.0","core_doi":"10.13026/nmdc-na68","version_doi":"10.13026/nqnw-7d62","is_latest_version":true,"short_description":"Annotations of ten types of medical decisions from discharge summaries in the MIMIC-III database.","publish_date":"2024-10-16","title":"MedDec: Medical Decisions for Discharge Summaries in the MIMIC-III Database","abstract":"Medical decisions directly impact individuals’ health and well-being. Extracting decision spans from clinical notes plays a crucial role in understanding medical decision-making processes. The MedDec dataset includes expert-annotated decisions of ten types, including defining problems, setting treatment goals, making drug-related decisions, performing therapeutic procedures, giving advice and precautions, and deferment decisions, spanning across eleven diverse patient phenotypes such as heart disease, lung disease, cancer, psychiatric disorders, and chronic pain, from 451 expert-annotated discharge summaries in the MIMIC-III database, where annotators identified decision spans within each discharge summary, noting character start and end positions alongside the decision type. The dataset enables the task of medical decision extraction, aiming to jointly extract and classify different types of medical decisions within clinical notes. This dataset also provides insights into medical decisions across diverse demographics, including gender, race, and English proficiency, sourced from the MIMIC-III database
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":9757610,"compressed_storage_size":2175853,"resource_type":"Database","access_policy":"Credentialed","topics":["medical decisions","natural language processing","span classification","discharge summary","mimic"],"source_url":"https://physionet.org/content/meddec/1.0.0/"},{"public_project_uuid":"364150d0-efe4-4e4e-ad8b-7efd82a2213c","slug":"race-ethnicity-clinical-text","version":"1.0.0","core_doi":"10.13026/t9ka-6k29","version_doi":"10.13026/z8tq-v658","is_latest_version":true,"short_description":"Two sets of gold-standard annotations for race and ethnicity information from clinical notes in MIMIC-III. Contains race and ethnicity label assignments and related information such as country of origin and spoken language.","publish_date":"2024-10-21","title":"C-REACT: Contextualized Race and Ethnicity Annotations for Clinical Text","abstract":"The Contextualized Race and Ethnicity Annotations for Clinical Text (C-REACT) dataset is a large publicly available corpus of sentences from clinical notes manually annotated for information related to race and ethnicity (RE). The corpus presented here contains 17,281 sentences drawn from 12,000 patients and their clinical notes at the Beth Israel Deaconess Medical Center critical care units between 2001 and 2012. This corpus contains two sets of reference standard annotations for RE data. The first set contains granular RE-information such as patient country of origin and spoken language. The second set of annotations contains RE labels manually assigned by clinicians. This corpus is intended to improve understanding about granular information related to RE contained within the clinical note and how this information might be used to infer RE.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":98940363,"compressed_storage_size":25330568,"resource_type":"Database","access_policy":"Credentialed","topics":["patient country information","race and ethnicity","patient language information","clinical notes"],"source_url":"https://physionet.org/content/race-ethnicity-clinical-text/1.0.0/"},{"public_project_uuid":"05dfa633-654d-40e8-a460-18043bf28f3d","slug":"dfci-cancer-outcomes-ehr","version":"1.0.0","core_doi":"10.13026/mjv2-8h15","version_doi":"10.13026/h2nj-p344","is_latest_version":true,"short_description":"The DFCI-imaging-student and DFCI-medonc-student AI models for extracting cancer outcomes from imaging reports and medical oncologist notes from electronic health records.","publish_date":"2024-10-24","title":"Shareable Artificial Intelligence to Extract Cancer Outcomes from Electronic Health Records for Precision Oncology Research","abstract":"Databases that link molecular data to clinical outcomes can inform precision cancer research into novel prognostic and predictive biomarkers. However, outside of clinical trials, cancer outcomes are typically recorded only in text form within electronic health records (EHRs). Artificial intelligence (AI) models have been trained to extract outcomes from individual EHRs. However, patient privacy restrictions have historically precluded dissemination of these models beyond the centers at which they were trained. In this study, the vulnerability of text classification models trained directly on protected health information to membership inference attacks was confirmed. A teacher-student distillation approach was applied to develop shareable models for annotating outcomes from imaging reports and medical oncologist notes. ‘Teacher’ models trained on EHR data from Dana-Farber Cancer Institute (DFCI) were used to label imaging reports and discharge summaries from the Medical Information Mart for Intensive Care (MIMIC)-IV dataset. ‘Student’ models were trained to use these MIMIC documents to predict the labels assigned by teacher models and sent to Memorial Sloan Kettering (MSK) for evaluation. The student models exhibited high discrimination across outcomes in both the DFCI and MSK test sets. These student models, “DFCI-imaging-student” and “DFCI-medonc-student,” are shared here.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1038013498,"compressed_storage_size":961999998,"resource_type":"Model","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/dfci-cancer-outcomes-ehr/1.0.0/"},{"public_project_uuid":"59b46b7d-11fd-40a9-ad31-7634e602b45f","slug":"covid-data-shared-learning","version":"1.0.0","core_doi":"10.13026/85z8-jq92","version_doi":"10.13026/1176-6c44","is_latest_version":true,"short_description":"COVID Data for Shared Learning (CDSL) is a multimodal database comprising de-identified structured health data and radiological images from 4,479 patients with COVID-19, as a comprehensive toolkit for developing predictive models.","publish_date":"2024-10-25","title":"COVID Data for Shared Learning (CDSL): A comprehensive, multimodal COVID-19 dataset from HM Hospitales","abstract":"COVID Data for Shared Learning (CDSL) is a multimodal database comprising de-identified medical data from 4,479 patients who were hospitalized with confirmed or suspected COVID-19 in the Spanish 'HM Hospitales' group from 2019-12-26 to 2021-02-13. The database provides tabular demographic, diagnostic, clinical and treatment information, as well as radiological images in JPG format, namely chest X-ray and computed tomography scans. The primary goal of CDSL is to develop a comprehensive toolkit to support researchers and institutions in building multimodal models for prediction, classification and optimization. CDSL database was promptly shared with the international research community at the onset of the COVID-19 pandemic to promote worldwide collaboration and, ultimately, guide policy decisions and facilitate effective response efforts studies on the disease.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":88432970474,"compressed_storage_size":86185401129,"resource_type":"Database","access_policy":"Contributor Review","topics":["multimodal database","covid-19","radiological images","open data","healthcare data","machine learning and ai"],"source_url":"https://physionet.org/content/covid-data-shared-learning/1.0.0/"},{"public_project_uuid":"2de49c5a-cc00-40aa-94b3-b47e7018e00f","slug":"chest-ct-sepsis-er","version":"1.0.0","core_doi":"10.13026/ct3n-e938","version_doi":"10.13026/zne5-qh18","is_latest_version":true,"short_description":"The database is intended to support a wide array of research studies involving radiomics in sepsis patients, helping to reduce barriers to the reproducibility of clinical research.","publish_date":"2024-10-28","title":"Chest Computed Tomography for patients with sepsis in the Emergency Department","abstract":"Sepsis is a systematic inflammatory response syndrome that can impact all vital organs. Lung is the most commonly involved organ that sepsis can cause lung injury. Lung injury can have a variety clinical presentations and can be captured by chest image studies. Computed tomography (CT) has a high resolution for lung parenchyma structure and provide valuable source of data to study structural changes of sepsis-induced lung injury. This dataset provide a high-resolution clinical tabular data and Chest CT for a total of 728 admissions with sepsis. Investigators from clinical medicine, computer vision and data scientist will find this database useful to study sepsis-induced lung injury.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":149514889386,"compressed_storage_size":90300659028,"resource_type":"Database","access_policy":"Contributor Review","topics":["sepsis"],"source_url":"https://physionet.org/content/chest-ct-sepsis-er/1.0.0/"},{"public_project_uuid":"9d7a22e0-7bf9-431f-a023-6d5a57a8bf81","slug":"mimic-iv-fhir","version":"2.1","core_doi":"10.13026/9hhp-ps57","version_doi":"10.13026/rrj1-ny66","is_latest_version":true,"short_description":"MIMIC-IV and MIMIC-IV-ED data mapped into FHIR resources.","publish_date":"2024-11-12","title":"MIMIC-IV on FHIR","abstract":"Fast Healthcare Interoperability Resources (FHIR) has emerged as a robust standard for healthcare data exchange. To explore the use of FHIR for the process of data harmonization, we converted the Medical Information Mart for Intensive Care IV (MIMIC-IV) and MIMIC-IV Emergency Department (MIMIC-IV-ED) databases into FHIR. We extended base FHIR to encode information in MIMIC-IV and aimed to retain the data in FHIR with minimal additional processing, aligning to US Core v4.0.0 where possible. A total of 24 profiles were created for MIMIC-IV data, and an additional 6 profiles were created for MIMIC-IV-ED data. Code systems and value sets were created from MIMIC terminology. We hope MIMIC-IV in FHIR provides a useful restructuring of the data to support applications around data harmonization, interoperability, and other areas of research.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":31387303667,"compressed_storage_size":31348909292,"resource_type":"Database","access_policy":"Credentialed","topics":["us core","mimic-iv","electronic health record","mimic","fast healthcare interoperability resources","fhir"],"source_url":"https://physionet.org/content/mimic-iv-fhir/2.1/"},{"public_project_uuid":"6d5fc5b3-a275-49ff-b876-21d34610c8d3","slug":"ms-cxr","version":"1.1.0","core_doi":"10.13026/ms5s-2a34","version_doi":"10.13026/9g2z-jg61","is_latest_version":true,"short_description":"MS-CXR is a new dataset containing 1162 chest X-ray bounding box labels paired with radiology text descriptions, annotated and verified by two board-certified radiologists.","publish_date":"2024-11-15","title":"MS-CXR: Making the Most of Text Semantics to Improve Biomedical Vision-Language Processing","abstract":"We release a new dataset, MS-CXR, with locally-aligned phrase grounding annotations by board-certified radiologists to facilitate the study of complex semantic modelling in biomedical vision–language processing. The MS-CXR dataset provides 1162 image–sentence pairs of bounding boxes and corresponding phrases, collected across eight different cardiopulmonary radiological findings, with an approximately equal number of pairs for each finding. This dataset complements the existing MIMIC-CXR v.2 dataset and comprises: 1. Reviewed and edited bounding boxes and phrases (1026 pairs of bounding box/sentence); and 2. Manual bounding box labels from scratch (136 pairs of bounding box/sentence).
\n\nThis large, well-balanced phrase grounding benchmark dataset contains carefully curated image regions annotated with descriptions of eight radiology findings, as verified by radiologists. Unlike existing chest X-ray benchmarks, this challenging phrase grounding task evaluates joint, local image-text reasoning while requiring real-world language understanding, e.g. to parse domain-specific location references, complex negations, and bias in reporting style. This data accompany work showing that principled textual semantic modelling can improve contrastive learning in self-supervised vision–language processing.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":3650322,"compressed_storage_size":411874,"resource_type":"Database","access_policy":"Credentialed","topics":["localization","phrase grounding","vision-language processing","chest x-ray"],"source_url":"https://physionet.org/content/ms-cxr/1.1.0/"},{"public_project_uuid":"1bc78cf5-cafa-4710-aa04-38438d12b5bd","slug":"nwicu-northwestern-icu","version":"0.1.0","core_doi":"10.13026/4je9-db46","version_doi":"10.13026/s84w-1829","is_latest_version":true,"short_description":"A freely available COVID-rich ICU database comprising de-identified health-related data from Northwestern Memorial Health Center (NHMC).","publish_date":"2024-11-19","title":"Northwestern ICU (NWICU) database","abstract":"Retrospective medical data collection is essential for advancing patient care, offering insights and supporting the development of health technology. The Medical Information Mart for Intensive Care (MIMIC)-III database has been instrumental in providing comprehensive critical care data to facilitate research efforts. Building upon the foundation of MIMIC-III, MIMIC-IV introduced a modular data structure and fostered an open-source community to enhance the data. Here we introduce Northwestern ICU database, a critical care dataset collected from a network of hospitals located in Chicago and harmonized with the MIMIC data structure.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":625830502,"compressed_storage_size":622528840,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/nwicu-northwestern-icu/0.1.0/"},{"public_project_uuid":"98743834-945b-4a33-a355-3afd2052e6ad","slug":"covidentify","version":"1.0.0","core_doi":"10.13026/fxah-ha58","version_doi":"10.13026/ncq1-vp79","is_latest_version":true,"short_description":"This contains wearable device data from Fitbit, Garmin, and Apple Watch users. The data is from April 2nd, 2020 to March 21st, 2021 and has been date-shifted. An appropriate amount has also shifted test dates for each user.","publish_date":"2024-11-25","title":"CovIdentify Dataset","abstract":"This dataset supports the study "A method for intelligent allocation of diagnostic testing by leveraging data from commercial wearable devices: a case study on COVID-19," which developed an Intelligent Testing Allocation (ITA) method. The study demonstrated the efficacy of using continuous digital biomarkers like resting heart rate and steps to enhance COVID-19 diagnostic testing positivity rates. The findings suggest significant potential for large-scale, symptom-independent surveillance testing to alleviate diagnostic test shortages. The provided data is from the CovIdentify study launched by Duke's BIG IDEAs Lab in the Biomedical Engineering Department. From April 2nd, 2020 to May 25th, 2021, 2,887 participants connected their smartwatches to the CovIdentify platform, including 1,689 Garmin, 1,091 Fitbit, and 107 Apple smartwatches
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":46204516,"compressed_storage_size":5507247,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/covidentify/1.0.0/"},{"public_project_uuid":"618652b2-6473-4bea-9f49-d6ca34f891e7","slug":"endoscapes-2023","version":"1.0.0","core_doi":"10.13026/qy7k-1v20","version_doi":"10.13026/czwq-jh81","is_latest_version":true,"short_description":"Endoscapes2023 enables the development of models for object detection, semantic and instance segmentation, and Critical View of Safety (CVS) prediction, contributing to safe laparoscopic cholecystectomy.","publish_date":"2024-12-11","title":"Endoscapes2023, A Critical View of Safety and Surgical Scene Segmentation Dataset for Laparoscopic Cholecystectomy","abstract":"Minimally invasive image-guided surgery heavily relies on vision. Deep learning models for surgical video analysis can support surgeons in visual tasks such as assessing the critical view of safety (CVS) in laparoscopic cholecystectomy, potentially contributing to surgical safety and efficiency. However, the performance, reliability, and reproducibility of such models are deeply dependent on the availability of data with high-quality annotations. To this end, we release Endoscapes2023, a dataset comprising 201 laparoscopic cholecystectomy videos with regularly spaced frames annotated with segmentation masks of surgical instruments and hepatocystic anatomy, as well as assessments of the criteria defining the CVS by three trained surgeons following a public protocol. Endoscapes2023 enables the development of models for object detection, semantic and instance segmentation, and CVS prediction, contributing to safe laparoscopic cholecystectomy.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":13543243858,"compressed_storage_size":12529046524,"resource_type":"Database","access_policy":"Restricted","topics":["surgical safety","computer assisted interventions","semantic segmentation","surgical data science","medical imaging analysis"],"source_url":"https://physionet.org/content/endoscapes-2023/1.0.0/"},{"public_project_uuid":"99df013c-d162-424b-82eb-be5e80771116","slug":"generated-codes-low-resource","version":"1.0.0","core_doi":"10.13026/bnc2-1a81","version_doi":"10.13026/09ng-2614","is_latest_version":true,"short_description":"9,606 Synthetic Discharge Summaries generated by GPT-3.5 based on combinations of ICD-10-code descriptions associated with real discharge summaries in MIMIC-IV. Focus on low resource codes.","publish_date":"2024-12-16","title":"MIMIC-IV-Ext-GPT-3_5-Generated-Discharge-Summaries-for-Low-Resource-Codes","abstract":"This dataset comprises 9,606 Synthetic Discharge Summaries generated by GPT-3.5 based on combinations of ICD-10-code descriptions associated with real discharge summaries in MIMIC-IV. As part of the generation process, GPT-3.5 was also tasked to code the discharge summaries (presented in square brackets in the Discharge sections) and predict a patient discharge status of 'DEAD' or 'ALIVE'.
\n\nEach generated discharge summary is presented with the following data – a list of ICD-10-CM and ICD-10-PCS codes upon which the prompt was based, the textual descriptions of said ICD-10-CM and ICD-10-PCS codes which are used in the prompt, the prompt itself, the raw output (including the predicted ICD-10 codes assigned to descriptions). We further provide the processed version of these discharge summaries we used in our augmentation experiments.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":151053953,"compressed_storage_size":21076517,"resource_type":"Database","access_policy":"Credentialed","topics":["icd coding","data augmentation","large language model"],"source_url":"https://physionet.org/content/generated-codes-low-resource/1.0.0/"},{"public_project_uuid":"7bb4d13d-80f4-4946-9fd8-c75a441cd9ab","slug":"senssmarttech","version":"1.0.0","core_doi":"10.13026/z1de-fn10","version_doi":"10.13026/fy9p-n277","is_latest_version":true,"short_description":"SensSmartTech is a unique multiparametric dataset recorded systematically at rest and during the relaxation after activity. It contains the simultaneously recorded electrocardiogram, phonocardiogram, arterial plethysmograms and seismocardiogram.","publish_date":"2024-12-19","title":"SensSmartTech database of cardiovascular signals synchronously recorded by an electrocardiograph, phonocardiograph, photoplethysmograph and accelerometer","abstract":"The SensSmartTech database comprises cardiovascular signals synchronously recorded by electrocardiograph (ECG), phonocardiograph (PCG), photoplethysmograph (PPG) and accelerometer (ACC) at heart rates at rest and after activity (HRs). It is composed of 338 30-s recordings taken from 32 healthy volunteers. Recordings contain information from 10 channels: 4 ECG (limb, V3 and V4 leads), 1 PCG (measured at heart apex), 4 PPG (measured at carotid and brachial arteries at two light wavelengths each) and 1 ACC (measured between V3 and V4 ECG electrodes), along with the recording time of each signal. The original recordings are stored in CSV format. Resampled recordings are stored in WFDB format. The large HR range with the average minimum on the set of 83 ± 11 bpm and the average maximum on the set of 143 ± 14 bpm, along with several recordings per heart relaxation period, enable studying of the cardiovascular dynamics and gaining insights into the corresponding cardio-respiratory and electro-mechanical couplings. Of particular clinical significance are the opportunities to derive corrections of multiparameter biomarkers for HR.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":812669619,"compressed_storage_size":277761540,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/senssmarttech/1.0.0/"},{"public_project_uuid":"d9a18dab-68e5-45ea-bf60-233049917f5a","slug":"hand-kinematics-semg","version":"1.0.0","core_doi":"10.13026/ns35-g888","version_doi":"10.13026/ch3e-c195","is_latest_version":true,"short_description":"A database of hand kinematics, high-density sEMG of forearm and wrist.","publish_date":"2025-01-17","title":"A database of hand kinematics, high-density sEMG of forearm and wrist for motion intent recognition","abstract":"Surface electromyography (sEMG) signals reflect spinal motor neuron activities and can be used as intuitive inputs for human-machine interaction (HMI) via movement intent recognition. The motor neuron potentials of far-field (wrist) and near-field (forearm) decomposed from high-density (HD) sEMG prospectively provide robust neural drives for HMI, which is a challenging research hotspot. However, there are no publicly available databases that include HD sEMG signals of forearm-wrist (FW) muscles, and hand kinematics (KIN). This paper presents the HD-FW KIN dataset that comprises HD 448-channel sEMG arrays distributed on the forearm and wrist with simultaneous recording of finger joint angles and finger flexion forces. This dataset contains muscle activities of 21 subjects while performing 20 hand gestures, and 9 individual or combined finger flexion under two force levels. The usability of HD-sEMG for hand gesture recognition, finger angle, and force prediction was validated. The proposed database allows a comprehensive extraction of the neural drive from the forearm and wrist, providing neural interfaces for the development of advanced prosthetic hand and wrist-worn consumer electronics.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":122483975458,"compressed_storage_size":90864215916,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/hand-kinematics-semg/1.0.0/"},{"public_project_uuid":"cce021a0-caf1-4704-8ba8-bb4030f11b64","slug":"b2ai-voice","version":"1.1","core_doi":"10.13026/37yb-1t42","version_doi":"10.13026/249v-w155","is_latest_version":false,"short_description":"A dataset of voice recordings and metadata to enable the development, benchmarking, and validation of clinically applicable machine-learning models for diagnosing a wide range of health conditions.","publish_date":"2025-01-17","title":"Bridge2AI-Voice: An ethically-sourced, diverse voice dataset linked to health information","abstract":"The human voice contains complex acoustic markers which have been linked to important health conditions including dementia, mood disorders, and cancer. When viewed as a biomarker, voice is a promising characteristic to measure as it is simple to collect, cost-effective, and has broad clinical utility. Recent advances in artificial intelligence have provided techniques to extract previously unknown prognostically useful information from dense data elements such as images. The Bridge2AI-Voice project seeks to create an ethically sourced flagship dataset to enable future research in artificial intelligence and support critical insights into the use of voice as a biomarker of health. Here we present Bridge2AI-Voice, a comprehensive collection of data derived from voice recordings with corresponding clinical information. Bridge2AI-Voice v1.0, the initial release, provides 12,523 recordings for 306 participants collected across five sites in North America. Participants were selected based on known conditions which manifest within the voice waveform including voice disorders, neurological disorders, mood disorders, and respiratory disorders. The initial release contains data considered low risk, including derivations such as spectrograms but not the original voice recordings. Detailed demographic, clinical, and validated questionnaire data are also made available.
","license":{"name":"Bridge2AI Voice Registered Access License"},"dua":{"name":"Bridge2AI Voice Registered Access Agreement"},"main_storage_size":4547685384,"compressed_storage_size":4525341024,"resource_type":"Database","access_policy":"Restricted","topics":["bridge2ai","voice"],"source_url":"https://physionet.org/content/b2ai-voice/1.1/"},{"public_project_uuid":"696dbe1b-ae56-4dfd-92d6-752141f35f35","slug":"nacef","version":"1.0.0","core_doi":"10.13026/14tj-m189","version_doi":"10.13026/m71c-9345","is_latest_version":false,"short_description":"Community-Acquired Pneumonia (CAP) poses a significant health risk, linked to high in-hospital morbidity and mortality rates. The dataset includes clinical details of 768 CAP patients at Clinica Universidad de La Sabana, Colombia.","publish_date":"2025-01-21","title":"Community-Acquired Pneumonia, Endotypes and Phenotypes (NACef): Prospective, observational cohort study of Translational Medicine","abstract":"Community-Acquired Pneumonia (CAP) remains a prominent infectious process associated with elevated in-hospital morbidity and mortality rates. Through the exploration of phenotypes, endotypes, and biomarkers, it becomes feasible to identify individuals at a heightened risk of adverse outcomes from CAP. The dataset encompasses clinical information from 768 patients diagnosed with CAP at Clinica Universidad de La Sabana, Colombia. Clinical data encompasses Baseline Clinical Data, In-hospital follow-up and data after hospital discharge. This repository has been utilized within the framework of a prospective research study, involving an observational cohort conducted in the domain of translational medicine. The dataset offers the opportunity for diverse statistical analyses, contributing to an enhanced comprehension of CAP, Endotypes, and Phenotypes. Moreover, this dataset proves valuable in the context of educational initiatives.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":263009,"compressed_storage_size":39996,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/nacef/1.0.0/"},{"public_project_uuid":"f08c1ffc-6f86-4739-a706-a5554fd48b8e","slug":"mimic-iv-ext-direct","version":"1.0.0","core_doi":"10.13026/2kfz-4798","version_doi":"10.13026/yf96-kc87","is_latest_version":true,"short_description":"A diagnostic reasoning dataset designed to evaluate the performance of large language models in aligning with human doctors when making diagnoses from clinical notes.","publish_date":"2025-01-21","title":"MIMIC-IV-Ext-DiReCT","abstract":"Large language models (LLMs) have recently demonstrated remarkable capabilities across a broad spectrum of tasks and applications, including the medical field. Models like GPT-4 excel in medical question answering but encounter challenges in interpretability when managing complex tasks in real clinical settings. To address this, we introduce the Diagnostic Reasoning Dataset for Clinical Notes (DiReCT), designed to evaluate the reasoning ability and interpretability of LLMs compared to human doctors. The dataset comprises 511 clinical notes (sourced from MIMIC-IV), each meticulously annotated by physicians, detailing the diagnostic reasoning process from initial observations to the final diagnosis.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":1253615,"compressed_storage_size":1220732,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/mimic-iv-ext-direct/1.0.0/"},{"public_project_uuid":"db2893d2-c343-46ea-a276-54b05e345c25","slug":"medical-cxr-vqa-dataset","version":"1.0.0","core_doi":"10.13026/y7r6-4j38","version_doi":"10.13026/1pm5-hy02","is_latest_version":true,"short_description":"Medical-CXR-VQA provides a large-scale LLM-enhanced dataset for visual question answering in medical chest x-ray images.","publish_date":"2025-01-21","title":"Medical-CXR-VQA dataset: A Large-Scale LLM-Enhanced Medical Dataset for Visual Question Answering on Chest X-Ray Images","abstract":"Medical Visual Question Answering (VQA) is an important task in medical multi-modal Large Language Models (LLMs), aiming to answer clinically relevant questions regarding input medical images. This technique has the potential to improve the efficiency of medical professionals while relieving the burden on the public health system, particularly in resource-poor countries. However, existing medical VQA datasets are small and only contain simple questions (equivalent to classification tasks), which lack semantic reasoning and clinical knowledge. Our previous work proposed a clinical knowledge-driven image difference VQA benchmark using a rule-based approach. However, given the same large-scale breadth of information coverage, the rule-based approach shows an 85% error rate on extracted labels. We trained an LLM method to extract labels with 62% increased accuracy. We also comprehensively evaluated our labels with 2 clinical experts on 100 samples to help us fine-tune the LLM. Based on the trained LLM model, we proposed a large-scale medical VQA dataset, Medical-CXR-VQA, derived from the MIMIC-CXR dataset and comprises 780,014 question-answer pairs, categorized into six types: abnormality (190,525 pairs), location (104,680 pairs), type (69,486 pairs), level (111,715 pairs), view (92,048 pairs), and presence (211,560 pairs).
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":297818508,"compressed_storage_size":40732373,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/medical-cxr-vqa-dataset/1.0.0/"},{"public_project_uuid":"b2948fb6-8fc8-414a-b182-5fe27e4e4ef1","slug":"ffa-ir-medical-report","version":"1.1.0","core_doi":"10.13026/k5rp-9h43","version_doi":"10.13026/f7w3-gm74","is_latest_version":true,"short_description":"Benchmark dataset for report generation based on fundus fluorescein angiography images and reports.","publish_date":"2025-01-21","title":"FFA-IR: Towards an Explainable and Reliable Medical Report Generation Benchmark","abstract":"Automatic medical report generation (MRG) towards describing life-threatening lesions from given medical images, such as Chest X-ray and Fundus Fluorescein Angiography (FFA), has been a long-standing research topic in machine learning and automatic medical diagnosis fields. However, existing MRG benchmarks only provide medical images and free-text reports without explainable annotations and reliable evaluation tools, hindering the current research advances from two aspects: First, existing methods can only predict reports without accurate explanation, undermining the trustworthiness of the diagnostic methods; Second, the comparison between predicted reports from MRG methods is unreliable based on the natural language generation (NLG) metrics.
\n\nTo address these issues, we propose an explainable and reliable MRG benchmark based on FFA Images and Reports (FFA-IR). Specifically, our FFA-IR dataset is featured from the following aspects: 1) Large-scale medical dataset. FFA-IR collects 766 reports along with 47,247 FFA images from clinical practice. 2) Explainable annotation. FFA-IR annotates 46 categories of lesions with a total of 12,166 regions. 3) Bilingual reports. FFA-IR provides both English and Chinese reports for each case. We hope that our FFA-IR can significantly advance research from both vision-and-language and medicine fields and improve the conventional retinal disease diagnosis procedure.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":13105215970,"compressed_storage_size":13101881873,"resource_type":"Database","access_policy":"Credentialed","topics":["fundus fluorescein angiography","medical report generation","vision and language","explainable and reliable evaluation"],"source_url":"https://physionet.org/content/ffa-ir-medical-report/1.1.0/"},{"public_project_uuid":"cb2291d3-90e1-4490-b578-429d0b3c1362","slug":"tame-pain","version":"1.0.0","core_doi":"10.13026/wp85-ea58","version_doi":"10.13026/20e2-1g10","is_latest_version":true,"short_description":"TAME Pain is a dataset that captures acoustic signals of pain and is augmented by annotating every sentence the participant speaks with details such as background and foreground noise, speech errors, and non-speech vocal features.","publish_date":"2025-01-21","title":"TAME Pain: Trustworthy AssessMEnt of Pain from Speech and Audio for the Empowerment of Patients","abstract":"Precise pain assessment is essential for medical professionals to provide appropriate treatment. However, not every patient can verbalize their pain due to various reasons, such as speech disorders or language barriers. In these cases, medical practitioners must rely on non-verbal signs to determine the pain level. The TAME Pain project aims to pave the way for the development of reliable pain assessment tools through advanced audio analysis.
\n\nWe aim to create a comprehensive dataset that captures acoustic signals to accurately predict pain levels. This dataset, approved by the University of Texas at Austin's institutional review board (IRB number: STUDY00004954), will enable the investigation of whether acoustic and non-acoustic signals extracted from healthy individuals subjected to pain can reliably indicate pain levels. We augment this dataset by annotating every single audio file, including every sentence spoken by the participant, with details such as background and foreground noise, speech errors, and non-speech vocal features. These annotations enable thorough audio analysis, facilitate pain studies, and aid in identifying both speech and non-speech pain cues. This dataset provides a resource for researchers and developers working on pain assessment technologies.
\n\nThe data collection and creation efforts are based at the University of Texas at Austin, with collaborative input from the University of Nottingham and the University of Southampton in the UK.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":473364086,"compressed_storage_size":470054488,"resource_type":"Database","access_policy":"Restricted","topics":["speech","audio","pain","cold pressor task"],"source_url":"https://physionet.org/content/tame-pain/1.0.0/"},{"public_project_uuid":"4022fdc3-27b4-4014-a6ff-8f8f9b459737","slug":"therlid","version":"1.0.0","core_doi":"10.13026/eskc-3e36","version_doi":"10.13026/tkww-8b64","is_latest_version":true,"short_description":"TherLiD is an open-source dataset of 13,251 paired temperature readings (contact and infrared) from MIMIC-IV and eICU databases. With added demographics and derived data, it supports research on racial and ethnic disparities in infrared thermometry.","publish_date":"2025-01-21","title":"TherLid: A Thermometry Linked Dataset","abstract":"A recent study showed that infrared (IR) sensors may be prone to calibration discrepancies among darker-pigmented patients. Similar disparities have already been verified in pulse oximetry. This raises questions about whether thermometry measurements may inadvertently overlook hypothermia, fever, or sepsis cases, potentially leading to delayed diagnoses and ultimately exacerbating poorer outcomes among vulnerable subpopulations.
\n\nTherLiD is a derived dataset from 3 Electronic Health Record databases: MIMIC-IV, eICU-CRD-1, and eICU-CRD-2. It consists of 13,251 temperature pairs, along with comprehensive demographic data and time-synchronized hospital information, offering a detailed profile of each patient. These pairs have one reference (contact thermometers - oral, core, and rectal) and one infrared-based (temporal) temperature value measured within a 1-hour time window, with temperature values between 30°C to 45°C. TherLiD not only provides high-quality, clinically relevant data but also offers a reproducible framework, allowing researchers to tailor the dataset to their specific research needs, including training machine learning models. This dataset was also built to facilitate temperature-related retrospective studies and promote research on racial and ethnic healthcare disparities.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":11206758,"compressed_storage_size":2471003,"resource_type":"Database","access_policy":"Credentialed","topics":["thermometry","intensive care unit","health equity","electronic health records"],"source_url":"https://physionet.org/content/therlid/1.0.0/"},{"public_project_uuid":"b44a6269-4479-4ad9-8213-dffb282f20dd","slug":"chexmask-cxr-segmentation-data","version":"1.0.0","core_doi":"10.13026/6eky-y831","version_doi":"10.13026/3705-zg36","is_latest_version":true,"short_description":"CheXmask Database is a 657,566 uniformly annotated chest radiographs with segmentation masks. Images were segmented using HybridGNet, with automatic quality control indicated by RCA scores.","publish_date":"2025-01-22","title":"CheXmask Database: a large-scale dataset of anatomical segmentation masks for chest x-ray images","abstract":"The CheXmask Database presents a comprehensive, uniformly annotated collection of chest radiographs, constructed from five public databases: ChestX-ray8, Chexpert, MIMIC-CXR-JPG, Padchest and VinDr-CXR. The database aggregates 657,566 anatomical segmentation masks derived from images which have been processed using the HybridGNet model to ensure consistent, high-quality segmentation. To confirm the quality of the segmentations, we include in this database individual Reverse Classification Accuracy (RCA) scores for each of the segmentation masks. This dataset is intended to catalyze further innovation and refinement in the field of semantic chest X-ray analysis, offering a significant resource for researchers in the medical imaging domain.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":40089510477,"compressed_storage_size":16784461855,"resource_type":"Database","access_policy":"Open","topics":["automatic quality assesment","chest x-ray segmentation","medical image segmentation"],"source_url":"https://physionet.org/content/chexmask-cxr-segmentation-data/1.0.0/"},{"public_project_uuid":"435c3054-7c24-414a-ad69-afd429c9f301","slug":"wearable-device-dataset","version":"1.0.0","core_doi":"10.13026/67z9-jh94","version_doi":"10.13026/zzf8-xv61","is_latest_version":false,"short_description":"Physiological signals(Electrodermal Activity,Blood Volume Pulse, Heart Rate, Temperature,etc) from 36 healthy volunteers collected during structured acute stress induction and aerobic/anaerobic exercise sessions using the Empatica E4 wearable device.","publish_date":"2025-01-24","title":"Wearable Device Dataset from Induced Stress and Structured Exercise Sessions","abstract":"This original dataset contains physiological signals collected during structured acute stress induction and aerobic and anaerobic exercise sessions using a wearable device. Blood volume pulse, motion-based activity, skin temperature, and electrodermal activity were recorded with the Empatica E4, a research-grade wearable. The stress induction protocol involved math and emotional tasks designed to provoke stress responses, interleaved with rest periods. Self-reported stress levels were also recorded during this procedure. For the exercise sessions, distinct routines on a stationary bike were created for aerobic and anaerobic activities. The dataset includes records from 36 healthy volunteers for stress sessions, 30 for aerobic exercise, and 31 for anaerobic exercise. By examining the variations in physiological signals, the effects of these activities can be analyzed. This dataset is a valuable resource for research on stress and exercise detection and classification.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":259375714,"compressed_storage_size":73121143,"resource_type":"Database","access_policy":"Open","topics":["wearable","anaerobic","aerobic","stress","exercise"],"source_url":"https://physionet.org/content/wearable-device-dataset/1.0.0/"},{"public_project_uuid":"939dde12-eb5c-48fb-9b2a-5e452da0b61c","slug":"llava-rad-mimic-cxr-annotation","version":"1.0.0","core_doi":"10.13026/6ey5-df78","version_doi":"10.13026/4ma4-k740","is_latest_version":true,"short_description":"This dataset provides GPT-4 extracted sections of radiology reports from MIMIC-CXR, complementing rule-based section extractions with additional reports with findings, and removing references to priors from findings.","publish_date":"2025-01-24","title":"LLaVA-Rad MIMIC-CXR Annotations","abstract":"LLaVA-Rad MIMIC-CXR features more accurate section extractions from MIMIC-CXR free-text radiology reports. Traditionally, rule-based methods were used to extract sections such as the reason for exam, findings, and impression. However, these approaches often fail due to inconsistencies in report structure and clinical language. In this work, we leverage GPT-4 to extract these sections more reliably, adding 237,073 image-text pairs to the training split and 1,952 pairs to the validation split. This enhancement afforded the development and fine-tuning of LLaVA-Rad, a multimodal large language model (LLM) tailored for radiology applications, achieving improved performance on report generation tasks.
\n\nThis resource is provided to support reproducibility and for the benefit of the research community, enabling further exploration in vision–language modeling. For more details, please refer to the accompanying paper [1].
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":868423093,"compressed_storage_size":65694683,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/llava-rad-mimic-cxr-annotation/1.0.0/"},{"public_project_uuid":"f84e336c-c6c3-4c16-8d43-9b380b360cef","slug":"music-sudden-cardiac-death","version":"1.0.1","core_doi":"10.13026/fa8p-he52","version_doi":"10.13026/z3m7-rf58","is_latest_version":true,"short_description":"The MUSIC study is a prospective, multicentre, longitudinal study designed to assess risk predictors of cardiac mortality and sudden cardiac death in ambulatory patients with chronic heart failure.","publish_date":"2025-01-24","title":"MUSIC (Sudden Cardiac Death in Chronic Heart Failure)","abstract":"The MUSIC (MUerte Subita en Insuficiencia Cardiaca) study is a prospective, multicentre, longitudinal study designed to assess risk predictors of cardiac mortality and sudden cardiac death (SCD) in ambulatory patients with chronic heart failure (CHF).
\n\nThe study population consisted of 992 patients with CHF consecutively enrolled from the specialized HF clinics of eight University Spanish Hospitals between April 2003 and December 2004, and followed up for a median of 44 months (until November 2008). All patients had a 3-lead resting electrocardiogram (ECG), a 24 h, 2-(4%) or 3-lead (96%) Holter ECG, chest X-ray, echocardiography, and blood laboratory parameters performed at enrolment.
\n\nPrimary outcomes were cardiac death, either sudden cardiac death (SCD) or pump failure death (PFD) at the end of the follow-up period.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":96890951588,"compressed_storage_size":46665996217,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/music-sudden-cardiac-death/1.0.1/"},{"public_project_uuid":"1fc62b8c-af4e-48af-9052-bfcb729b0de8","slug":"script-x2b8-dataset","version":"1.0.0","core_doi":"10.13026/av6f-8r24","version_doi":"10.13026/235w-zn26","is_latest_version":true,"short_description":"This dataset contains electronic health record (EHR) data from ICU patients receiving mechanical ventilation, aggregated on a daily basis, along with annotations of intubation, extubation, tracheostomy days, and cases of failed extubation. Data can b","publish_date":"2025-01-28","title":"SCRIPT X2B8 Dataset: per-day clinical features to model successful next-day extubation","abstract":"Criteria to identify patients who are ready to be liberated from mechanical ventilation are imprecise, often resulting in prolonged mechanical ventilation or reintubation, both of which are associated with adverse outcomes. We sought to determine whether machine learning applied to the electronic health record could predict extubation success. With the X2B8 dataset, we provide cleaned next-day extubation labels for an internal cohort of 696 patients and 9,828 ICU days, and an external test cohort of 333 patients and 2,835 ICU days. With an eye towards future deployment of a model that could be used during daily clinical rounds, we aggregated data from 37 clinical features from midnight to 8 AM. Data have been deidentified per Health Insurance Portability and Accountability Act (HIPAA) Safe Harbor rules. We use this dataset in a manuscript examining different machine learning models to predict successful extubation (see Usage section for details) and share it for others to work with.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":2903090,"compressed_storage_size":767951,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/script-x2b8-dataset/1.0.0/"},{"public_project_uuid":"0104fb95-b655-4b36-aedf-c3333e14f078","slug":"minute-level-step-count-nhanes","version":"1.0.0","core_doi":"10.13026/5htm-7515","version_doi":"10.13026/ah0j-3z47","is_latest_version":false,"short_description":"Minute level step counts obtained from five step counting algorithms for raw accelerometry data, and minute level Activity Counts, MIMS, wear predictions, and wear flags for all participants who wore accelerometers in NHANES 2011-2014.","publish_date":"2025-01-28","title":"Minute level step counts and physical activity data from the National Health and Nutrition Examination Survey (NHANES) 2011-2014","abstract":"The National Health and Nutrition Examination Survey (NHANES) is a nationally representative study that collects demographic, socioeconomic, dietary, and health-related information from 10,000 Americans annually. In Wave G (2011-2012) and Wave H (2013-2014), participants wore an ActiGraph GT3X+ on the non-dominant wrist for seven consecutive days. Publicly available data include minute-level wear predictions (wake wear, sleep wear, unknown, nonwear), data quality flags, and monitor-independent movement summary units (MIMS). Additionally, raw triaxial accelerometer data recorded at 80 Hz are provided, enabling detailed analyses beyond MIMS.
\n\nWe applied five step counting algorithms and the ActiGraph Activity Count (AC) algorithm to the raw data to create a dataset of minute-level step counts. The dataset also includes minute level AC, MIMS, wear predictions, and wear flags for all participants who wore accelerometers in NHANES 2011–2014. These data facilitate detailed investigations of accelerometry-measured physical activity patterns in a nationally representative U.S. population. Potential analyses include examination of within-day activity patterns and their association with health outcomes.
","license":{"name":"Creative Commons Zero 1.0 Universal Public Domain Dedication"},"dua":null,"main_storage_size":2513577818,"compressed_storage_size":2513395619,"resource_type":"Database","access_policy":"Open","topics":["nhanes","accelerometry","steps","physical activity"],"source_url":"https://physionet.org/content/minute-level-step-count-nhanes/1.0.0/"},{"public_project_uuid":"a1b40359-df9e-4b2f-b809-6ccef3785eac","slug":"cgmacros","version":"1.0.0","core_doi":"10.13026/8mak-rs10","version_doi":"10.13026/3z8q-x658","is_latest_version":true,"short_description":"CGMacros contains information from two continuous glucose monitors (CGM), food macronutrients, food photographs, physical activity, and anonymized participant demographics, anthropometric measurements and health parameters.","publish_date":"2025-01-28","title":"CGMacros: a scientific dataset for personalized nutrition and diet monitoring","abstract":"We present CGMacros, a dataset containing multimodal information from two continuous glucose monitors (CGM), food macronutrients, food photographs, and physical activity, in addition to anonymized participant demographics, anthropometric measurements and health parameters from blood analyses and gut microbiome profiles. CGMacros contains data from 45 study participants (15 healthy adults, 16 with pre-diabetes, and 14 with Type 2 diabetes) who consumed meals with varying and known macronutrient compositions in a free-living setting for ten consecutive days. To our knowledge, this is the first database of its kind to be made publicly available. CGMacros, and larger publicly available datasets that we hope may follow, are essential to democratize academic research in personalized nutrition and algorithmic approaches to automated diet monitoring.
","license":{"name":"Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":658404462,"compressed_storage_size":657529467,"resource_type":"Database","access_policy":"Open","topics":["diabetes","continuous glucose monitors","obesity","machine learning","postprandial glucose response","food macronutrients","metabolic models","food photographs","personalized nutrition"],"source_url":"https://physionet.org/content/cgmacros/1.0.0/"},{"public_project_uuid":"97dad99f-5d45-4e3c-9087-acff385801c7","slug":"vqa-evaluation-mimic-cxr","version":"1.0.0","core_doi":"10.13026/tz5h-1w39","version_doi":"10.13026/cvsk-ny21","is_latest_version":true,"short_description":"This dataset provides 224 VQAs for 40 test set cases, and 111 VQAs for 23 validation set cases of the MIMIC CXR dataset.","publish_date":"2025-01-28","title":"Visual Question Answering evaluation dataset for MIMIC CXR","abstract":"MIMIC CXR [1] is a large publicly available dataset of chest radiographs in DICOM format with free-text radiology reports. In addition, labels for the presence of 12 different chest-related pathologies, as well as of any support devices, and overall normal/abnormal status were made available via the MIMIC Chest X-ray JPG (MIMIC-CXR-JPG) [2] labels, which were generated using the CheXpert and NegBio algorithms.
\n\nBased on these labels, we created a visual question answering dataset comprising 224 questions for 48 cases from the official test set, and 111 questions for 23 validation cases. A majority (68%) of the questions are close-ended (answerable with yes or no), and focus on the presence of one out of 15 chest pathologies, or any support device, or generically on any abnormality, whereas the remaining open-ended questions inquire about the location, size, severity or type of a pathology/device, if present in the specific case, indicated by the MIMIC-CXR-JPG labels.
\n\nFor each question and case we also provide a reference answer, which was authored by a board-certified radiologist (with 17 years of post-residency experience) based on the chest X-ray and original radiology report
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":177623,"compressed_storage_size":114265,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/vqa-evaluation-mimic-cxr/1.0.0/"},{"public_project_uuid":"d773f7bf-f6c7-4d17-9b2f-7cac181dcc32","slug":"symile-mimic","version":"1.0.0","core_doi":"10.13026/762h-qf80","version_doi":"10.13026/3vvj-s428","is_latest_version":true,"short_description":"A multimodal clinical dataset consisting of CXRs, ECGs, and blood labs, designed to evaluate Symile, a simple contrastive loss that accommodates any number of modalities and allows any model to produce representations for each modality.","publish_date":"2025-01-28","title":"Symile-MIMIC: a multimodal clinical dataset of chest X-rays, electrocardiograms, and blood labs from MIMIC-IV","abstract":"Symile-MIMIC is a multimodal clinical dataset derived from MIMIC-IV and MIMIC-CXR, consisting of chest X-rays (CXRs), electrocardiograms (ECGs), and blood laboratory tests. It was developed to evaluate Symile, a contrastive learning objective designed to handle multiple modalities and enable any model to generate representations for each modality. The dataset explores whether ECG and blood work collected at admission are predictive of a CXR taken shortly thereafter. Symile-MIMIC includes 11,622 admissions, split into training, validation, and test sets with no patient overlap between splits. Each sample contains an ECG, a CXR, and up to 50 common blood lab results. This module provides: (1) the dataset in CSV format, (2) pre-processed tensors of the dataset, (3) the code to generate the dataset from MIMIC-IV and MIMIC-CXR, and (4) the best model checkpoint trained on the Symile-MIMIC dataset using the Symile objective.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":34439695796,"compressed_storage_size":8129800256,"resource_type":"Database","access_policy":"Credentialed","topics":["contrastive learning","model","cxr","multimodal","chest x-ray","mimic","ecg","electrocardiogram","database"],"source_url":"https://physionet.org/content/symile-mimic/1.0.0/"},{"public_project_uuid":"8ac70b45-4572-4ea3-bc1c-b06fef60f55f","slug":"cxrgraph","version":"1.0.0","core_doi":"10.13026/f2sv-v647","version_doi":"10.13026/p7kf-t860","is_latest_version":true,"short_description":"CXRGraph is a structured radiology report dataset built upon RadGraph and tailored for the Automatic Radiology Report Generation task. It can identify more task-relevant information such as abnormalities and hallucinated prior references.","publish_date":"2025-02-03","title":"CXRGraph: Using Information Extraction to Normalize the Training Data for Automatic Radiology Report Generation","abstract":"CXRGraph is a dataset of structured radiology reports dataset following the RadGraph format, which has been tailored for the Automatic Radiology Report Generation (ARRG) task. CXRGraph assorts clinical information from full-text radiology reports into five entity types and four relation types similar to RadGraph. CXRGraph introduces three entity attributes, which are optionally associated with an entity to provide additional information (e.g. abnormality) and handle hallucinated prior references for the ARRG task. We manually annotated the reports originally formatted in RadGraph, including 550 MIMIC-CXR reports for model training and evaluation and 50 CheXpert reports for evaluating the model generalization ability. By using the ground-truth data, we developed a joint entity and relation model, achieving a micro-F1 of 96.6% and 96.1% on named entity recognition, 94.0% and 89.8% on entity attribute recognition, and 89.5% and 86.6% on relation extraction, on the MIMIC-CXR and CheXpert test sets, respectively. Using the trained model, we automatically annotated 227,835 MIMIC-CXR reports. Both the ground-truth and inference data are available in CXRGraph. Given that the MIMIC-CXR and RadGraph have been de-identified already, no protected health information (PHI) is included.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":630760439,"compressed_storage_size":100574950,"resource_type":"Database","access_policy":"Restricted","topics":["natural language processing","relation extraction","named entity recognition","information extraction","structured radiology report"],"source_url":"https://physionet.org/content/cxrgraph/1.0.0/"},{"public_project_uuid":"b578d397-10be-4c81-b970-902af226f460","slug":"medical-diff-vqa","version":"1.0.1","core_doi":"10.13026/hbsg-tk87","version_doi":"10.13026/e6dd-cn74","is_latest_version":true,"short_description":"MIMIC-Diff-VQA provides a large-scale dataset for Difference visual question answering in medical chest x-ray images.","publish_date":"2025-02-03","title":"Medical-Diff-VQA: A Large-Scale Medical Dataset for Difference Visual Question Answering on Chest X-Ray Images","abstract":"The task of Difference Visual Question Answering involves answering questions about the difference between a pair of main and reference images. This process is consistent with the radiologist's diagnosis practice that compares the current image with the reference before concluding the report. We've assembled a new dataset, called Medical-Diff-VQA, for this purpose. Unlike previous medical VQA datasets, ours is the first one designed specifically for the Difference Visual Question Answering task, with questions crafted to suit the Assessment-Diagnosis-Intervention-Evaluation treatment procedure employed by medical professionals. The Medical-Diff-VQA dataset, a derivative of the MIMIC-CXR dataset, consists of questions categorized into seven categories: abnormality (145,421), location (84,193), type (27,478), level (67,296), view (56,265), presence (155,726), and difference(164,324). The 'difference' questions are specifically for comparing two images. In total, the Medical-Diff-VQA dataset contains 700,703 question-answer pairs derived from 164,324 pairs of main and reference images.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":311733683,"compressed_storage_size":21894282,"resource_type":"Database","access_policy":"Credentialed","topics":["vqa","chest x-ray","difference visual question answering","difference vqa","visual question answering"],"source_url":"https://physionet.org/content/medical-diff-vqa/1.0.1/"},{"public_project_uuid":"f01b47c7-f6c3-44a3-9f73-28de76cb9c71","slug":"synthetic-mention-corpora","version":"1.0.0","core_doi":"10.13026/e0kj-js78","version_doi":"10.13026/p5pn-ty93","is_latest_version":true,"short_description":"We present the Synthetic Mention Corpora for Disease Entity Recognition and Normalization, containing 128000 disease mentions from the UMLS disorder group, generated by an LLM. This corpus aims to improve these tasks in biomedical and clinical texts.","publish_date":"2025-02-03","title":"Synthetic Mention Corpora for Disease Entity Recognition and Normalization","abstract":"Named Entity Recognition (NER) and Entity Normalization (EN) are fundamental tasks in information extraction, particularly in the biomedical and clinical domains. NER identifies textual mentions of entities, while EN maps these mentions to unique identifiers within a structured vocabulary. However, the biomedical domain presents unique challenges for NER, including the diverse and inconsistent lexical representations of biomedical concepts, such as non-standard terminology, abbreviations, complex phrases, and frequent misspellings in clinical texts. Additionally, rare entities are often underrepresented in training datasets and may lack detailed descriptions or synonyms in knowledge graphs, limiting the quality of training data for Disease Entity Recognition (DER) and Disease Entity Normalization (DEN). To address this, we present the Synthetic Mention Corpora for Disease Entity Recognition and Normalization, a dataset comprising 128,000 synthetic disease mentions generated using a fine-tuned LLaMa-2-13B-Chat model. These mentions are derived from the Unified Medical Language System (UMLS) disorder group. This corpus aims to enhance the development of more robust systems for disease entity identification and linking in biomedical and clinical text, addressing current limitations in training data availability.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":97534962,"compressed_storage_size":20548822,"resource_type":"Database","access_policy":"Open","topics":["named entity recognition","data augmentation","entity normalization","machine learning","nlp"],"source_url":"https://physionet.org/content/synthetic-mention-corpora/1.0.0/"},{"public_project_uuid":"d8b61099-aece-4b38-b36e-f547179b89fc","slug":"labelled-notes-hospital-course","version":"1.2.0","core_doi":"10.13026/fh2q-4148","version_doi":"10.13026/5gte-bv70","is_latest_version":true,"short_description":"This dataset presents a collection of preprocessed and labeled clinical notes derived from \"MIMIC-IV-Note\", and aims to facilitate the development of ML models focused on summarizing brief hospital courses (BHC) from clinical notes.","publish_date":"2025-02-03","title":"MIMIC-IV-Ext-BHC: Labeled Clinical Notes Dataset for Hospital Course Summarization","abstract":"This dataset presents a curated collection of preprocessed and labeled clinical notes derived from the MIMIC-IV-Note database. The primary aim of this resource is to facilitate the development and training of machine learning models focused on summarizing brief hospital courses (BHC) from clinical discharge notes.
\n\nThe dataset contains 270,033 meticulously cleaned and standardized clinical notes containing an average token length of 2,267, ensuring usability for machine learning (ML) applications. Each clinical note is paired with a corresponding BHC summary, providing a robust foundation for supervised learning tasks. The preprocessing pipeline employed uses regular expressions to address common issues in the raw clinical text, such as special characters, extraneous whitespace, inconsistent formatting, and irrelevant text, to produce a high-quality, structured dataset with separated clinical note sections through appropriate headings.
\n\nBy offering this resource, we aim to support healthcare professionals and researchers in their efforts to enhance patient care through the automation of BHC summarization. This dataset is ideal for exploring various NLP techniques, developing predictive models, and improving the efficiency and accuracy of clinical documentation practices. We invite the research community to utilize this dataset to advance the field of medical informatics and contribute to better health outcomes.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":2726467690,"compressed_storage_size":863062775,"resource_type":"Database","access_policy":"Credentialed","topics":["natural language processing","brief hospital course","text summarization","machine learning","clinical notes"],"source_url":"https://physionet.org/content/labelled-notes-hospital-course/1.2.0/"},{"public_project_uuid":"023bf002-7109-4f30-a376-300e5b1e50d1","slug":"med-palm2-mimic-cxr","version":"1.0.0","core_doi":"10.13026/a8e5-yx97","version_doi":"10.13026/7wmp-jx90","is_latest_version":true,"short_description":"This work further refines the labels associated with CheXpert in MIMIC-CXR-JPG 2.0.0 by filtering with Med-PaLM 2 followed by verification by manual review by three US board-certified radiologists.","publish_date":"2025-02-04","title":"Application of Med-PaLM 2 in the refinement of MIMIC-CXR labels","abstract":"MIMIC-CXR is a large, open source dataset that is widely-used in medical AI research. One of the limitations of this dataset is the lack of ground truth labels for the chest X-ray studies. Prior work has extracted structured labels from the MIMIC-CXR radiology report text using CheXpert, a natural language processing (NLP) model. As comprehensive expert validation of these labels is cost-prohibitive, there exists a need for scalable methods of identifying NLP-derived labels that would benefit from manual review. We have developed prompts for extraction of clinically-relevant labels using a clinically-trained large language model, Med-PaLM 2, which we selectively applied to MIMIC-CXR radiology reports. A subset of cases where the Med-PaLM 2 results differed from the previously published CheXpert labels were reviewed by three US board certified radiologists to establish a ground truth. Of these differing labels, Med-PaLM 2 achieved an accuracy of 66%, compared to 19% of CheXpert. Our results demonstrate the potential use of medically-oriented large language models such as Med-PaLM 2 in both label extraction and identifying cases for manual review. This dataset offers 1,378 radiologist-verified ground truth labels to the MIMIC-CXR project.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":395576,"compressed_storage_size":39324,"resource_type":"Database","access_policy":"Restricted","topics":["mimic-cxr labels"],"source_url":"https://physionet.org/content/med-palm2-mimic-cxr/1.0.0/"},{"public_project_uuid":"c11692f6-fbe5-4d78-bed6-9f69b13a2de9","slug":"latte-cxr","version":"1.0.0","core_doi":"10.13026/0c10-gm86","version_doi":"10.13026/0pw2-je90","is_latest_version":true,"short_description":"This dataset includes bounding box-statement pairs for chest X-ray images, derived from radiologists’ eye-tracking data (for explainability) and annotations, for local visual-language models.","publish_date":"2025-02-04","title":"LATTE-CXR: Locally Aligned TexT and imagE, Explainable dataset for Chest X-Rays","abstract":"Local annotation of medical data is both expensive and time-consuming due to the high cost of expert annotators, the precision required for accurate annotation, and the inherent challenges of medical diagnosis. To address these problems, we developed LATTE-CXR, a chest X-ray dataset with locally aligned image-text pairs, derived from the REFLACX dataset. LATTE-CXR supports tasks requiring local image-text annotations, such as phrase grounding, caption-guided object detection, and image captioning with region-level descriptions. By extracting statements from radiology reports corresponding to REFLACX annotated abnormalities, this dataset includes 3926 bounding box-statement pairs (with repeating statements) from 1668 MIMIC-CXR image readings in the REFLACX dataset. Additionally, we automatically generated 13751 bounding box-sentence pairs from 2,742 chest X-ray readings, utilizing timestamped eye-tracking data and transcribed reports from REFLACX. The eye-tracking bounding boxes are linked to corresponding annotated bounding boxes if they share a sentence, providing a comprehensive framework for assessing model explainability.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":4896529,"compressed_storage_size":2180946,"resource_type":"Database","access_policy":"Restricted","topics":["chest x-ray dataset","eye-tracking","automatically generated dataset","caption-guided object detection","localization","image captioning with region-level description","grounded radiology report generation","phrase grounding","xai","multi-modal learning","local visual-language models"],"source_url":"https://physionet.org/content/latte-cxr/1.0.0/"},{"public_project_uuid":"441e420d-a0ac-439d-8d57-620d7328d442","slug":"dreamt","version":"2.0.0","core_doi":"10.13026/dztc-dv77","version_doi":"10.13026/0vrv-nn81","is_latest_version":false,"short_description":"Dataset for Real-time sleep stage EstimAtion using Multisensor wearable Technology","publish_date":"2025-02-05","title":"DREAMT: Dataset for Real-time sleep stage EstimAtion using Multisensor wearable Technology","abstract":"Sleep is an intrinsic part of human life, and recent advancements in wearable technology and machine learning have promised continuous and non-invasive methods of tracking sleep health and patterns, providing an important facet to a more holistic understanding of well-being. However, it is still challenging to achieve consistent and reliable real-time estimates of sleep stages using only smartwatches. This is especially true for individuals with irregular sleep patterns or sleep disorders. A major contributing factor is the distinct lack of publicly accessible, large-scale datasets that allow researchers and engineers to validate their wearable sleep staging algorithms against a population with diverse sleep patterns. Here, we present DREAMT, Dataset for Real-time sleep stage EstimAtion using Multisensor wearable Technology, a new dataset collected from 100 participants, which includes high-resolution signals from a smartwatch, expert sleep technician-annotated sleep stage labels, and clinical metadata related to sleep health and disorders.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":161545924027,"compressed_storage_size":26517544599,"resource_type":"Database","access_policy":"Restricted","topics":["biomedical","time series classification","wearable","sleep disorders"],"source_url":"https://physionet.org/content/dreamt/2.0.0/"},{"public_project_uuid":"c81ed534-00d3-4d40-a6ef-9f30b211b3aa","slug":"openox-repo","version":"1.1.0","core_doi":"10.13026/e821-1858","version_doi":"10.13026/yq6c-z215","is_latest_version":false,"short_description":"A repository of matched arterial oxygen and pulse oximeter readings obtained under controlled conditions, with high-frequency physiologic waveforms and skin color measurements.","publish_date":"2025-02-19","title":"OpenOximetry Repository","abstract":"The OpenOximetry Repository is a structured database designed to store clinical and laboratory pulse oximetry data and allows for consolidation of data sets held by collaborating organizations. Matched or independent readings of oxygen saturations, arterial blood gas samples, high frequency processed and unprocessed waveforms, and other physiologic metadata are supported (e.g. skin color readings). Contribution of data to the repository is encouraged; a data dictionary and standardized data collection protocols are provided to ensure consistent data archival and interpretation. The goal of this dataset is to enable researchers to interrogate the effects of physiologic and device variables on pulse oximeter performance.
\n\nA set of controlled desaturation studies is included in the initial release with ongoing updates planned as more data are available from both controlled lab desaturation studies and real world prospective clinical trials. Data captured during these studies may include paired pulse oximeter and arterial blood gas readings throughout a range of oxygen saturations, patient demographics and comorbidities, longitudinal quantitative and qualitative skin color measurement (using multiple methods at multiple anatomic locations), finger diameter, continuous ECG, arterial blood pressure, EtCO2/EtO2, unprocessed photoplethysmography (PPG), pH, PaCO2, hemoglobin, met-Hgb, CO-Hgb, respiratory rate and multiple other variables depending on the data source.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":5298776343,"compressed_storage_size":3301474801,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/openox-repo/1.1.0/"},{"public_project_uuid":"7c2ced95-4753-4c6d-a842-4c193c2b0e79","slug":"multimodal-nback-music","version":"1.0.0","core_doi":"10.13026/gr11-3d06","version_doi":"10.13026/6vh4-dk68","is_latest_version":true,"short_description":"A multimodal dataset containing fNIRS data along with a wide range of physiological signals like EDA, HR, PPG, etc over the course of n-back experiments in presence of music.","publish_date":"2025-02-26","title":"A Multimodal Dataset for Investigating Working Memory in Presence of Music","abstract":"We present the accompanying dataset to the study "A Multimodal Dataset for Investigating Working Memory in Presenceof Music". The experiment is conducted with the aim of investigating the viability of music as an intervention to regulate cognitive arousal and performance states. We recorded the multimodal physiological signals and behavioral data during a working memory task called the n-back task while the background music was playing. We requested the participants to provide the music, and two types of music were employed with the calming and exciting content. The calming music was played during the first session of the experiment, and the exciting music was presented during the second session. Each session includes an equal number of 1-back and 3-back task blocks, where 22 trials are presentedwithin each task block. A total number of 16 task blocks are implemented in each session (8 blocks of 1-back task and 8 blocks of 3-back task). In this experiment,11 participants/subjects originally participated, while we removed participants/subjects with small modalities. The recorded signals are skin conductance (SC), electrocardiogram (ECG), skin surface temperature (SKT), respiration, photoplethysmography (PPG), functional near-infrared spectroscopy (fNIRS), electromyogram (EMG), de-identified facial expression scores, sequence of correct/incorrect responses, and reaction time.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":1706486245,"compressed_storage_size":316409245,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/multimodal-nback-music/1.0.0/"},{"public_project_uuid":"0de3aa0e-304f-4c9a-8570-09f19f45a558","slug":"archehr-qa-bionlp-task-2025","version":"1.0","core_doi":"10.13026/zzax-sy62","version_doi":"10.13026/p17h-kg38","is_latest_version":false,"short_description":"A dataset for grounded question answering (QA) from electronic health records (EHRs).","publish_date":"2025-02-27","title":"ArchEHR-QA: BioNLP at ACL 2025 Shared Task on Grounded Electronic Health Record Question Answering","abstract":"Responding to patients’ medical inbox messages through patient portals is increasingly a contributor to clinician burden. To this end, automatically generating answers to questions from patients considering their medical records is important. The overarching goal of the ArchEHR-QA 2025 shared task is to develop automated responses to patients' questions by generating answers that are grounded in key clinical evidence from their electronic health records (EHRs). The proposed dataset, ArchEHR-QA, comprises hand-curated, realistic patient questions (reflective of patient portal messages), relevant focus areas identified within these questions (as determined by a clinician), corresponding clinician-rewritten versions (crafted to aid in formulating responses), and note excerpts providing essential clinical context. The task is to construct coherent answers to input questions that must be grounded in the provided clinical note excerpts.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":205607,"compressed_storage_size":34259,"resource_type":"Challenge","access_policy":"Credentialed","topics":["electronic health record","patient portals","clinicians","question answering"],"source_url":"https://physionet.org/content/archehr-qa-bionlp-task-2025/1.0/"},{"public_project_uuid":"fd4af2f1-100e-48e7-85fd-be432fed427c","slug":"pifir","version":"1.0.0","core_doi":"10.13026/e0cx-1k53","version_doi":"10.13026/d51v-j343","is_latest_version":true,"short_description":"A corpus of PET-CT reports annotated for terminology relevant to fungal infections. Ideal for validation of named entity recognition and relation extraction methods.","publish_date":"2025-02-27","title":"PIFIR: PET-CT Invasive Fungal Infection Reports","abstract":"Surveillance of invasive fungal infection (IFI) in clinical settings is a laborious process requiring a detailed review of patient medical history. One of the key sources of clinical information is imaging reports: radiologist-produced free-text reports summarising findings and observations from a scan. Positron emission tomography combined with computed tomography (PET-CT) is a medical imaging modality particularly useful in ruling out IFIs and evaluating the response to anti-fungal therapy. The data was generated to facilitate the development of an automated tool for the detection of IFI. The PET-CT Invasive Fungal Infection Reports (PIFIR) corpus contains 201 de-identified reports annotated by radiologists for terminology suggestive of the presence of IFI. These include IFI-related concepts and certainty cues, as well as relations between them. We release the annotation schema and the original reports along with the corresponding annotation files. We anticipate the PIFIR corpus to be useful in the development and validation of named entity recognition and relation extraction methods with a focus on clinical data. Such methods can be instrumental in processing other types of clinical documentation (clinical notes, nursing notes) with various downstream tasks in mind.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":817116,"compressed_storage_size":530755,"resource_type":"Database","access_policy":"Credentialed","topics":["clinical documentation","invasive fungal infections","nlp","information extraction"],"source_url":"https://physionet.org/content/pifir/1.0.0/"},{"public_project_uuid":"77ef8252-0bac-44fd-bda7-565d53ae23d2","slug":"archehr-qa-bionlp-task-2025","version":"1.1","core_doi":"10.13026/zzax-sy62","version_doi":"10.13026/f9xr-rr81","is_latest_version":false,"short_description":"A dataset for grounded question answering (QA) from electronic health records (EHRs).","publish_date":"2025-02-28","title":"ArchEHR-QA: BioNLP at ACL 2025 Shared Task on Grounded Electronic Health Record Question Answering","abstract":"Responding to patients’ medical inbox messages through patient portals is increasingly a contributor to clinician burden. To this end, automatically generating answers to questions from patients considering their medical records is important. The overarching goal of the ArchEHR-QA 2025 shared task is to develop automated responses to patients' questions by generating answers that are grounded in key clinical evidence from their electronic health records (EHRs). The proposed dataset, ArchEHR-QA, comprises hand-curated, realistic patient questions (reflective of patient portal messages), relevant focus areas identified within these questions (as determined by a clinician), corresponding clinician-rewritten versions (crafted to aid in formulating responses), and note excerpts providing essential clinical context. The task is to construct coherent answers to input questions that must be grounded in the provided clinical note excerpts.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":202227,"compressed_storage_size":34287,"resource_type":"Challenge","access_policy":"Credentialed","topics":["clinicians","question answering","electronic health record","patient portals"],"source_url":"https://physionet.org/content/archehr-qa-bionlp-task-2025/1.1/"},{"public_project_uuid":"c142089c-217c-41c9-ab97-9e1b51636032","slug":"openox-repo","version":"1.1.1","core_doi":"10.13026/e821-1858","version_doi":"10.13026/be2e-cn29","is_latest_version":true,"short_description":"A repository of matched arterial oxygen and pulse oximeter readings obtained under controlled conditions, with high-frequency physiologic waveforms and skin color measurements.","publish_date":"2025-02-28","title":"OpenOximetry Repository","abstract":"The OpenOximetry Repository is a structured database designed to store clinical and laboratory pulse oximetry data and allows for consolidation of data sets held by collaborating organizations. Matched or independent readings of oxygen saturations, arterial blood gas samples, high frequency processed and unprocessed waveforms, and other physiologic metadata are supported (e.g. skin color readings). Contribution of data to the repository is encouraged; a data dictionary and standardized data collection protocols are provided to ensure consistent data archival and interpretation. The goal of this dataset is to enable researchers to interrogate the effects of physiologic and device variables on pulse oximeter performance.
\nA set of controlled desaturation studies is included in the initial release with ongoing updates planned as more data are available from both controlled lab desaturation studies and real world prospective clinical trials. Data captured during these studies may include paired pulse oximeter and arterial blood gas readings throughout a range of oxygen saturations, patient demographics and comorbidities, longitudinal quantitative and qualitative skin color measurement (using multiple methods at multiple anatomic locations), finger diameter, continuous ECG, arterial blood pressure, EtCO2/EtO2, unprocessed photoplethysmography (PPG), pH, PaCO2, hemoglobin, met-Hgb, CO-Hgb, respiratory rate and multiple other variables depending on the data source.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":5982707503,"compressed_storage_size":3415890822,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/openox-repo/1.1.1/"},{"public_project_uuid":"f9db9878-4c45-44bd-9855-ae331baa7509","slug":"mc-med","version":"1.0.0","core_doi":"10.13026/xgx1-7x47","version_doi":"10.13026/jz99-4j81","is_latest_version":false,"short_description":"A multimodal dataset of deidentified clinical and physiological data from emergency department visits, aimed at enabling research on patient outcomes, care processes, and the impact of continuous monitoring on treatment during and after the COVID-19.","publish_date":"2025-03-03","title":"Multimodal Clinical Monitoring in the Emergency Department (MC-MED)","abstract":"Emergency department (ED) patients often present with undiagnosed complaints, and can exhibit rapidly evolving physiology. Therefore, data from continuous physiologic monitoring, in addition to the electronic health record, is essential to understand the acute course of illness and responses to interventions. The complexity of ED care and the large amount of unstructured multimodal data it produces has limited the accessibility of detailed ED data for research. We release Multimodal Clinical Monitoring in the Emergency Department (MC-MED), a comprehensive, multimodal, and de-identified clinical and physiological dataset. MC-MED includes 118,385 adult ED visits to an academic medical center from 2020 to 2022. Data include continuously monitored vital signs, physiologic waveforms (electrocardiogram, photoplethysmogram, respiration), patient demographics, medical histories, orders, medication administrations, laboratory and imaging results, and visit outcomes. MC-MED is the first dataset to combine detailed physiologic monitoring with clinical events and outcomes for a large, diverse ED population.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":515385765850,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mc-med/1.0.0/"},{"public_project_uuid":"03e93a38-0434-4c0d-bffa-01c5990c9e29","slug":"cardiac-implantable-device-cxr","version":"1.0.0","core_doi":"10.13026/jwdf-q907","version_doi":"10.13026/swv0-cd11","is_latest_version":true,"short_description":"This dataset comprises 11,094 converted DICOM and smartphone images of Cardiac Implantable Electronic Devices (CIEDs), collected from 897 patients. It aims to facilitate the development of algorithms for CIED detection and classification.","publish_date":"2025-03-04","title":"Dataset for Segmentation and Classification of Cardiac Implantable Electronic Devices in Chest X-Rays","abstract":"This dataset encompasses a comprehensive collection of 2,321 chest radiographs (originally DICOM, converted to PNG) and 11,072 smartphone images documenting various cardiac implantable electronic devices (CIEDs) such as implantable pacemakers, cardioverter defibrillators, cardiac resynchronization therapy devices, and cardiac monitors, collected from 896 patients at the Charité - Universitätsmedizin Berlin over a decade (January 2012 to January 2022). The dataset's primary objective is to advance the development and validation of automated techniques for precise segmentation and classification of CIEDs across both traditional and smartphone-generated imagery. It includes a diverse array of 27 CIED models from four manufacturers (Medtronic, Biotronic, Boston Scientific, Abbott Laboratories including St. Jude Medical), captured in anterior-posterior or posterior-anterior chest radiographs using five different smartphone brands. This dataset aims to underpin the creation of deep learning models for accurately identifying CIED types, manufacturers, and models. Designed to be a robust foundation for research and clinical applications, it facilitates the exploration of innovative machine learning solutions for CIED identification leveraging both original radiogrpahs and smartphone images, thereby addressing a critical need in cardiac care technology.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":48616227554,"compressed_storage_size":37456592726,"resource_type":"Database","access_policy":"Restricted","topics":["medical imaging","cardiac implantable electronic devices","chest x-ray","radiology"],"source_url":"https://physionet.org/content/cardiac-implantable-device-cxr/1.0.0/"},{"public_project_uuid":"147c6c90-374c-4dfc-83dd-6fb468b820c6","slug":"mietic","version":"1.0.0","core_doi":"10.13026/nf30-0z23","version_doi":"10.13026/q1nc-2e47","is_latest_version":true,"short_description":"MIMIC-IV-Ext Triage Instruction Corpus includes 9,629 ED triage cases organized by the five-level ESI, enabling LLMs to improve triage accuracy. It provides CSV data, generation prompts, expert validation samples, and SQL QC scripts.","publish_date":"2025-03-04","title":"MIMIC-IV-Ext Triage Instruction Corpus","abstract":"Emergency department (ED) overcrowding leads to delayed care, increased patient risk, and inefficient resource use. The MIMIC-IV-Ext Triage Instruction Corpus (MIETIC) addresses this by providing 9,629 structured triage cases from MIMIC-IV, aligned with the Emergency Severity Index (ESI).
\nMIETIC supports large language model (LLM) training for AI-assisted triage, improving accuracy, consistency, and risk assessment. The dataset includes chief complaints, vital signs, demographics, and medical history, ensuring realistic triage decision-making. Developed through automated quality control and expert validation, MIETIC enhances model performance in high-risk and moderate-risk classification.
\nAvailable in CSV formats, MIETIC enables research in clinical NLP, AI-driven triage, and decision-support tools. The dataset module includes:
\nMIETIC provides a standardized, reproducible dataset to advance AI-driven emergency triage, optimizing accuracy, efficiency, and resource allocation.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":23815677,"compressed_storage_size":4470612,"resource_type":"Database","access_policy":"Credentialed","topics":["large language models","machine learning","nlp","clinical decision support","emergency severity index","emergency triage"],"source_url":"https://physionet.org/content/mietic/1.0.0/"},{"public_project_uuid":"b56abec3-336d-4d4b-861c-a792b1595a9f","slug":"micro-nav","version":"1.0.0","core_doi":"10.13026/tyj5-dd44","version_doi":"10.13026/nx6s-hf22","is_latest_version":false,"short_description":"This study explores the respiratory microbiome's role in nosocomial lower respiratory tract infections in ICU patients. Conducted in Chía, Colombia, revealing the microbiome's impact on disease progression.","publish_date":"2025-03-10","title":"Microbiological, Immunological and Biochemical Characteristics of the Development of Ventilator Associated Pneumonia","abstract":"The respiratory microbiome plays a critical role in metabolism, immune system maturation, and protection against pathogens. Traditionally, respiratory microbiology in pneumonia focused on identifying a specific pathogen, often disregarding normal oral flora as contaminants. However, recent research highlights the importance of the pulmonary microbiome's altered composition (dysbiosis) and its association with diseases such as Ventilator-Associated Pneumonia (VAP). The relationship between dysbiosis and immune dysregulation remains unclear, raising questions about whether dysbiosis is a cause or consequence of disease progression. This dataset explores the respiratory microbiome's role in nosocomial lower respiratory tract infections in ICU patients.
\n\nThis prospective, observational dataset was built in an ICU in Chia, Colombia, focusing on adult patients requiring mechanical ventilation without initial pneumonia. Clinical data and biological samples, including bronchoalveolar lavage, orotracheal secretion, and blood, were collected at multiple time points. Microbiome analysis involved DNA extraction, gene sequencing of the 16S ribosomal unit, and bioinformatic analysis to characterize microbial diversity. Concurrently, pulmonary, and systemic inflammatory markers were quantified using ELISA.
\n\nThe dataset enrolled 141 patients, generating a comprehensive dataset on demographic, clinical, and microbiome variables. This research not only provides insights into the microbial ecology of VAP but also establishes a repository of data for future studies. The findings underscore the complex interplay between the microbiome and disease in critically ill patients, with implications for improving clinical outcomes in the ICU setting.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":5075530,"compressed_storage_size":396589,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/micro-nav/1.0.0/"},{"public_project_uuid":"4282d246-0f4d-49c2-9ca7-b8b5679a935a","slug":"medvh","version":"1.0.0","core_doi":"10.13026/wjx8-sx53","version_doi":"10.13026/8ymd-c338","is_latest_version":false,"short_description":"MedVH provides a visual hallucination evaluation benchmark for large language models in the medical context. It formulates tests using chest X-ray images, including multi-choice question answering and long-text generation tasks.","publish_date":"2025-03-11","title":"MedVH: Towards Systematic Evaluation of Hallucination for Large Vision Language Models in the Medical Context","abstract":"Large Vision Language Models (LVLMs) have recently achieved superior performance in various tasks on natural image and text data, which inspires a large amount of studies for LVLMs fine-tuning and training. Despite their advancements, there has been scant research on the robustness of these models against hallucination when fine-tuned on smaller datasets. In this study, we introduce a new benchmark dataset, the Medical Visual Hallucination evaluation benchmark (MedVH), to evaluate the hallucination of domain-specific LVLMs. MedVH comprises six tasks to evaluate hallucinations in LVLMs within the medical context, which includes tasks for a comprehensive understanding of textual and visual input, as well as long textual response generation.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1475671088,"compressed_storage_size":1470005250,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/medvh/1.0.0/"},{"public_project_uuid":"e6fd01a1-f7df-4a6e-9ffd-610d1efaa65a","slug":"leipzig-heart-center-ecg","version":"1.0.0","core_doi":"10.13026/2ha4-km43","version_doi":"10.13026/7a4j-vn37","is_latest_version":true,"short_description":"This annotated ECG database for paediatric and CHD patients features 12-lead and intracardiac recordings, supporting advanced diagnostic algorithms.","publish_date":"2025-03-19","title":"Leipzig Heart Center ECG-Database: Arrhythmias in Children and Patients with Congenital Heart Disease","abstract":"Interpretation of Electrocardiograms (ECG) is increasingly complemented by algorithms. These algorithms are based on large datasets. This ECG database consists of children and adults with congenital heart defects (CHD) including many arrhythmia annotations. This dataset, notable for its manual annotations and inclusion of intracardiac electrograms alongside traditional 12-lead ECGs, offers 1075.85 minutes of recordings (in total 113924 annotated beats) that capture various cardiac rhythms and arrhythmias such as supraventricular tachycardia and ventricular tachycardia. The data were meticulously collected from patients undergoing electrophysiological studies, with subsequent annotations by expert reviewers using the LightWAVE® software. The significance of this database lies in its focus on pediatric arrhythmias and arrhythmias of patients with congenital heart defect areas currently underrepresented in existing datasets, which predominantly feature adult pathologies. This resource aims to enhance algorithmic development for ECG interpretation, leveraging machine learning to improve diagnosis and treatment outcomes in these sensitive groups. The dataset not only serves as a critical tool for developing precision medicine but also sets a precedent for future expansions to include a broader spectrum of congenital heart defects conditions, thereby supporting the evolution of cardiac care through advanced computational techniques.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":3576526560,"compressed_storage_size":1310633104,"resource_type":"Database","access_policy":"Open","topics":["arrhythmias","artificial intelligence","chd","12-lead","intracardiac recordings","annotated","congenital heart disease","ecg"],"source_url":"https://physionet.org/content/leipzig-heart-center-ecg/1.0.0/"},{"public_project_uuid":"8ac99b82-a9a7-4f0c-8922-f3358e90ad2d","slug":"rexerr-v1","version":"1.0.0","core_doi":"10.13026/7ytb-x609","version_doi":"10.13026/9dns-vd94","is_latest_version":true,"short_description":"Chest X-Ray reports containing synthetic errors based upon the MIMIC-CXR database. Errors were injected using LLMs and sampled across common human and AI model errors.","publish_date":"2025-03-19","title":"ReXErr-v1: Clinically Meaningful Chest X-Ray Report Errors Derived from MIMIC-CXR","abstract":"Interpreting medical images and writing radiology reports is a critical yet challenging task in healthcare. Despite their importance, both human-written and AI-generated reports are liable to errors, leaving a need for robust and representative datasets that capture the diversity of errors present across different mediums of report generation. Thus, we present Chest X-Ray Report Errors (ReXErr-v1), a new dataset based on MIMIC-CXR and constructed using large language models (LLMs) that contains synthetic error reports for the majority of MIMIC-CXR. Developed with input from board-certified radiologists, ReXErr-v1 contains plausible errors that closely mimic those found in real-world scenarios. Furthermore, ReXErr-v1 utilizes a novel sampling methodology that selects three errors to inject among a set of frequent errors made by both human and AI models. We include errors both at report and sentence level, improving the versatility of ReXErr-v1. Our dataset can enhance future AI reporting tools by aiding the development and evaluation of report-generation and error-screening algorithms.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":533669659,"compressed_storage_size":86749857,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/rexerr-v1/1.0.0/"},{"public_project_uuid":"f89c4b1f-e06d-4394-b08e-e1761a221d70","slug":"ehrcon-consistency-of-notes","version":"1.0.1","core_doi":"10.13026/x1ea-np24","version_doi":"10.13026/m4vd-y789","is_latest_version":true,"short_description":"Dataset for Checking Consistency between Unstructured Notes and Structured Tables in Electronic Health Records","publish_date":"2025-03-19","title":"EHRCon: Dataset for Checking Consistency between Unstructured Notes and Structured Tables in Electronic Health Records","abstract":"Electronic Health Records (EHRs) are integral for storing comprehensive patient medical records, combining structured data (e.g., medications) with detailed clinical notes (e.g., physician notes). These elements are essential for straightforward data retrieval and provide deep, contextual insights into patient care. However, they often suffer from discrepancies due to unintuitive EHR system designs and human errors, posing serious risks to patient safety. To address this, we developed EHRCon, a new dataset and task specifically designed to ensure data consistency between structured tables and unstructured notes in EHRs. EHRCon was crafted in collaboration with healthcare professionals using the MIMIC-III EHR dataset, and includes manual annotations of 4,101 entities across 105 clinical notes checked against database entries for consistency. EHRCon has two versions, one using the original MIMIC-III schema, and another using the OMOP CDM schema, in order to increase its applicability and generalizability.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":8327081,"compressed_storage_size":1550140,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/ehrcon-consistency-of-notes/1.0.1/"},{"public_project_uuid":"558f6a22-58bd-4cd3-914e-e324ed79747f","slug":"cxr-lt-iccv-workshop-cvamd","version":"2.0.0","core_doi":"10.13026/5mbq-9k80","version_doi":"10.13026/ryj9-x506","is_latest_version":true,"short_description":"CXR-LT 2024 was a challenge for long-tailed, multi-label, and zero-shot thorax disease classification on chest X-rays, held at MICCAI 2024. This page contains long-tailed labels for 45 diseases from the CXR-LT 2024 and 2023 challenges.","publish_date":"2025-03-19","title":"CXR-LT: Multi-Label Long-Tailed Classification on Chest X-Rays","abstract":"Chest radiography presents a \"long-tailed\" distribution of findings, where a few diseases are common, but most are rare. Diagnosis is further complicated by its multi-label nature, as patients often exhibit multiple co-occurring findings. While recent research has attempted to address the long-tailed medical image classification problem, the interplay between class imbalance and label co-occurrence remains underexplored. The CXR-LT 2024 challenge builds on the success of CXR-LT 2023, expanding the dataset of 377,110 chest X-rays (CXRs) to 45 disease labels, including 19 new rare disease findings. This year’s challenge introduces three tasks: (i) long-tailed classification on a large, noisy test set, (ii) long-tailed classification on a manually annotated \"gold standard\" subset, and (iii) zero-shot generalization to five previously unseen disease findings. CXR-LT 2024 addresses critical challenges in long-tailed, multi-label, and zero-shot learning for medical imaging by synthesizing state-of-the-art solutions from the international research community. Further, our dataset contributions — expanding disease coverage to better reflect real-world clinical settings — offer a valuable resource for future research. This project contains labels from the CXR-LT 2024 and CXR-LT 2023 challenges, as well as a related subset used in the MICCAI 2023 paper, \"How Does Pruning Impact Multi-Label Long-Tailed Learning?\"
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":394345899,"compressed_storage_size":94561870,"resource_type":"Challenge","access_policy":"Credentialed","topics":["disease classification","computer-aided diagnosis","artificial intelligence","chest x-ray","deep learning","long-tailed learning","cardiopulmonary disease","zero-shot learning"],"source_url":"https://physionet.org/content/cxr-lt-iccv-workshop-cvamd/2.0.0/"},{"public_project_uuid":"18db2d5e-4d6b-40cc-9391-89fb45ba738f","slug":"alott","version":"1.0.0","core_doi":"10.13026/46e5-hh23","version_doi":"10.13026/sbq5-dy17","is_latest_version":true,"short_description":"Carescape is a deidentified 9 month period of high resolution telemetry data linked to Electronic Medical Records.","publish_date":"2025-03-19","title":"ALarms, Outcomes Telemetry with Timing (ALOTT): a Bedside-EMR Database","abstract":"ALOTT is a pilot project that gathered telemetry data from 270 beds and over 15,000 hospital admissions from September 2018 through November 2020 at The James Cancer Hospital and Ross Heart Hospital. ALOTT contains telemetry waveforms, such as electrocardiogram and blood oxygen, at 60-240Hz and temperature, pulse, and perfusion measurements at two-second intervals linked to Electronic Medical Record (EMR) data. These EMR data include patient admissions, demographics, diagnosis, history, orders, labs, medications, allergies, and nurse flowsheet chart events, including vitals, risk scores, ventilator use, and emergency response team (ERT) events. This dataset was constructed to facilitate the creation of algorithms to predict ERT events.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":2573036114409,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/alott/1.0.0/"},{"public_project_uuid":"da0d576f-8028-4bb5-9044-406e2e9e2820","slug":"mimic-iv-ext-cekg","version":"1.0.0","core_doi":"10.13026/wqvp-h188","version_doi":"10.13026/qr9d-6t52","is_latest_version":true,"short_description":"The MIMIC-IV-Ext-CEKG dataset is crafted for object-centric process mining in healthcare, specifically to create clinical event knowledge graphs for patients with multimorbidity, as well as for data mining and machine learning tasks.","publish_date":"2025-04-09","title":"MIMIC-IV-Ext-CEKG: A Process-Oriented Dataset Derived from MIMIC-IV for Enhanced Clinical Insights","abstract":"Maintaining a healthy population is essential for improving quality of life and overall societal well-being. One approach to achieving a healthy population is by improving patients' care pathways. This is particularly vital for patients with multiple chronic conditions, who require well-coordinated care across various medical specialties. One approach to improving and analyzing the care pathways of these types of patients is process mining. The clinical event knowledge graph is a recent framework in process mining introduced for patients with multimorbidity that facilitates standardized interpretation of care pathways by linking to ICD-10 and SNOMED-CT. It also facilitates storing recent multi-entity event data in the event graph and analyzing care pathways for multimorbid patients from multiple perspectives. MIMIC-IV is a dataset that facilitates data analysis in healthcare; however, it is not specialized for process mining and requires extensive data preprocessing to prepare it for process mining. This paper contributes to the MIMIC-IV-Ext-CEKG dataset, an extracted dataset from MIMIC-IV that facilitates using the Clinical Event Knowledge Graph framework and other process mining tasks. This paper describes its characteristics and how it is extracted from the MIMIC-IV dataset. MIMIC-IV-Ext-CEKG facilitates deploying MIMIC-IV for process mining.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":57885915420,"compressed_storage_size":57822283517,"resource_type":"Database","access_policy":"Credentialed","topics":["process mining","multi entity process mining","mimic","object centric event log","clinical event knowledge graph"],"source_url":"https://physionet.org/content/mimic-iv-ext-cekg/1.0.0/"},{"public_project_uuid":"58802e8b-fe77-4d3d-abb2-6b8ab7cad5ac","slug":"mimic-iii-ext-tpatchgnn","version":"1.0.0","core_doi":"10.13026/rzea-ah40","version_doi":"10.13026/ckn0-3868","is_latest_version":true,"short_description":"The processed MIMIC-III dataset for the benchmark of Irregular Multivariate Time Series Forecasting: A Transformable Patching Graph Neural Networks Approach.","publish_date":"2025-04-09","title":"MIMIC-III-Ext-tPatchGNN","abstract":"This dataset is a curated subset of MIMIC-III (v1.4), specifically formatted to facilitate reproducibility of the experiments in the work t-PatchGNN. It serves as part of a benchmark designed for forecasting irregular multivariate clinical time series, that is, given a set of historical Irregular Multivariate Time Series (IMTS) observations and forecasting queries, the forecasting problem aims to accurately forecast the values in correspondence to these queries. This requires addressing key challenges such as missing data, variable sampling rates, and complex temporal dependencies. The dataset includes patient records with diverse physiological measurements, each sampled at irregular intervals, reflecting real-world clinical scenarios. It is structured to capture both short-term and long-term temporal patterns, making it well-suited for evaluating machine learning models in medical time series forecasting. By providing a standardized benchmark, this dataset aims to advance research in predictive modeling for healthcare, enabling the development of robust algorithms that can handle irregular and sparse clinical data. The dataset’s applications extend to critical areas such as early disease detection, patient risk stratification, and treatment outcome prediction, making it a valuable resource for the medical AI and machine learning communities.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":872044252,"compressed_storage_size":39959752,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimic-iii-ext-tpatchgnn/1.0.0/"},{"public_project_uuid":"0f2619ef-4395-4d0b-b102-538d2a3f5992","slug":"mimic-iii-ext-verifact-bhc","version":"1.0.0","core_doi":"10.13026/5cn9-ga59","version_doi":"10.13026/abat-g475","is_latest_version":false,"short_description":"A clinician-labeled dataset for fact-checking long-form clinical text against patient EHRs. The dataset contains LLM-written and human-written Brief Hospital Course summaries decomposed to atomic claim and sentence propositions with annotations.","publish_date":"2025-04-09","title":"MIMIC-III-Ext-VeriFact-BHC: Labeled Propositions From Brief Hospital Course Summaries for Long-form Clinical Text Evaluation","abstract":"The VeriFact-BHC dataset is designed to verify the factuality of long-form text written about a patient against their own electronic health record. There is increasing interest in using large language models (LLMs) to generate clinical text in patient care applications, yet this text needs to be evaluated for factual errors and hallucinations prior to committing text to a patient’s permanent medical record. Text written about a patient should be internally consistent with information already known about the patient, such as that stored in their medical records.
\nVeriFact-BHC contains long-form Brief Hospital Course (BHC) clinical narratives typically found in a discharge summary that have been decomposed into text proposition statements. From 100 patients in the MIMIC-III Clinical Database v1.4, we consider two types of BHC text: a human-written BHC and a LLM-generated BHC. The original human clinician-written BHC is extracted from the discharge summary note. The LLM-generated BHC is composed by a LLM using the patient’s longitudinal clinical notes from the hospital admission. Each BHC is decomposed in two ways: sentence propositions and atomic claim propositions. The remaining electronic health record (EHR) notes for each patient serves as a patient-specific reference of facts that is used by clinicians and VeriFact to assign labels. A total of 13,070 propositions are annotated by multiple clinicians with a ground truth established via majority voting and manual adjudication. Also provided are labels assigned by the VeriFact artificial intelligence system and labels assessing whether propositions are valid from a first-order logic standpoint. The reference EHR for each patient is provided in both machine-readable and PDF formats.
\nBy offering this dataset, we hope to spur further investigation and creation of computational systems for automatic chart review and patient-specific fact verification. We invite the research community to utilize this dataset to develop better methods to guardrail patient-specific LLM-generated clinical text.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":419346071,"compressed_storage_size":419099137,"resource_type":"Database","access_policy":"Credentialed","topics":["long-form text","chart review","artificial intelligence","clinical notes","text reranking","brief hospital course","natural language processing","atomic claim","hybrid retrieval","clinical informatics","clinical medicine","fact verification","retrieval-augmented generation","logical atomism","text embedding","formal logic","large language models","llm-as-a-judge","electronic health records","llm evaluation"],"source_url":"https://physionet.org/content/mimic-iii-ext-verifact-bhc/1.0.0/"},{"public_project_uuid":"8d7b4175-092d-48f4-80ce-1bc571309e80","slug":"ecg-capable-smartwatches","version":"1.0.0","core_doi":"10.13026/rs30-nm12","version_doi":"10.13026/7018-y383","is_latest_version":true,"short_description":"This database allows the study of the feasibility of using ECG-capable smartwatches as diagnostic tools, focusing on their compliance with clinical standards and their ability to measure critical ECG parameters beyond AF detection.","publish_date":"2025-04-09","title":"Electrocardiogram-Capable Smartwatches: Assessing Their Clinical Accuracy and Application","abstract":"Wearable technology has progressed significantly so that today's smartwatches are able to offer advanced health monitoring functions such as electrocardiogram (ECG). This study contains data that can be used to clinically assess the accuracy of four leading smartwatch models: Apple Watch Series 9, Samsung Galaxy Watch 6, Fitbit Sense 2, and Withings ScanWatch. Using the standardized protocol of the International Electrotechnical Commission (IEC 60601-2-25:2011) and a METRON PS-440 patient simulator, controlled ECG signals were generated and recorded to assess smartwatches performance against a reference electrocardiograph (Philips TC30). Parameters such as heart rate (30 to 300 beats per minute), R-wave amplitude (500 to 2000 µV), and ST-segment elevation/depression (-800 to +800 µV) were configured in the patient simulator and recorded 5 times in all devices. This extensive dataset provides valuable information on the feasibility of smartwatches as diagnostic tools, and both their strengths and limitations for clinical and remote cardiovascular monitoring can be observed.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":43076838,"compressed_storage_size":19312255,"resource_type":"Database","access_policy":"Restricted","topics":["ischemia","fitbit sense","ambulatory","samsung galaxy watch","st-segment","apple watch","withings scanwatch","smartwatch"],"source_url":"https://physionet.org/content/ecg-capable-smartwatches/1.0.0/"},{"public_project_uuid":"1d4ae932-f526-46d7-88be-eafd7be3fc28","slug":"archehr-qa-bionlp-task-2025","version":"1.2","core_doi":"10.13026/zzax-sy62","version_doi":"10.13026/nrhg-f267","is_latest_version":false,"short_description":"A dataset for grounded question answering (QA) from electronic health records (EHRs).","publish_date":"2025-04-11","title":"ArchEHR-QA: BioNLP at ACL 2025 Shared Task on Grounded Electronic Health Record Question Answering","abstract":"Responding to patients’ medical inbox messages through patient portals is increasingly a contributor to clinician burden. To this end, automatically generating answers to questions from patients considering their medical records is important. The overarching goal of the ArchEHR-QA 2025 shared task is to develop automated responses to patients' questions by generating answers that are grounded in key clinical evidence from their electronic health records (EHRs). The proposed dataset, ArchEHR-QA, comprises hand-curated, realistic patient questions (reflective of patient portal messages), relevant focus areas identified within these questions (as determined by a clinician), corresponding clinician-rewritten versions (crafted to aid in formulating responses), and note excerpts providing essential clinical context. The task is to construct coherent answers to input questions that must be grounded in the provided clinical note excerpts.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1084412,"compressed_storage_size":192611,"resource_type":"Challenge","access_policy":"Credentialed","topics":["electronic health record","clinicians","patient portals","question answering"],"source_url":"https://physionet.org/content/archehr-qa-bionlp-task-2025/1.2/"},{"public_project_uuid":"3709d29e-ed78-47d4-bd32-ac2553589061","slug":"aipatient-kg","version":"1.0.0","core_doi":"10.13026/jxcs-fq11","version_doi":"10.13026/vjrq-9328","is_latest_version":true,"short_description":"This project integrates MIMIC-III and CORAL electronic health records into knowledge graphs to improve medical analysis and enhance decision-making capabilities. Resources include two knowledge graph snapshots and two question-and-answering datasets.","publish_date":"2025-04-15","title":"AIPatient KG: MIMIC-III and CORAL Electronic Health Records based Patient Knowledge Graph","abstract":"This study integrates the MIMIC-III and CORAL electronic health records into knowledge graphs to enhance their utility for advanced medical analysis and decision-making. MIMIC-III contains comprehensive data from over 40,000 patients, while CORAL focuses on oncology-specific information from 40 patients, aiding in complex medical reasoning. We used a LLM (Large Language Model)-based Named Entity Recognition approach to extract relevant medical information from these datasets, independently verified by domain experts, and constructed the AIPatient and CORAL Knowledge Graph in Neo4j. This graph supports the AIPatient system, which simulates patient interactions for advanced decision support. Additionally, we introduce MIMIC-III and CORAL Question and Answering sets, which are created for evaluating system performance such as accuracy, robustness and stability.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1408577,"compressed_storage_size":127659,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/aipatient-kg/1.0.0/"},{"public_project_uuid":"afa767ad-5f3f-4e1e-ae17-3a2c52f5d9be","slug":"temporal-respiratory-support","version":"1.1.0","core_doi":"10.13026/4nm5-h710","version_doi":"10.13026/wewp-sj67","is_latest_version":true,"short_description":"A benchmark dataset offering hourly records over a 90-day period for 50,920 ICU subjects, including dynamic pulmonary function data and a spectrum of covariates for respiratory intervention analyses.","publish_date":"2025-04-15","title":"A Temporal Dataset for Respiratory Support in Critically Ill Patients","abstract":"We present a temporal benchmark dataset for clinical respiratory intervention tasks in intensive care unit (ICU) patients, derived from the MIMIC-IV v2.2 dataset. The data consists of 50,920 adult ICU patients and includes 90-day hourly ventilation data, laboratory results, vital signs, and details of treatment interventions such as respiratory support interventions, vasopressor administration, and continuous renal replacement therapy. Additionally, it encompasses static variables such as ICU length of stay, discharge outcome, and death outcome. While there are existing open ICU datasets, the limited availability of benchmark datasets designed for clinical treatments in respiratory support motivated the development of the data. The data facilitates various tasks including time-series analysis, survival analysis, sequential decision-making tasks, and reinforcement learning.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":25713453024,"compressed_storage_size":452546831,"resource_type":"Database","access_policy":"Credentialed","topics":["oberservational data","time-series"],"source_url":"https://physionet.org/content/temporal-respiratory-support/1.1.0/"},{"public_project_uuid":"14d34845-f3c1-4a96-bb22-63e8d53f4db2","slug":"shdb-af","version":"1.0.1","core_doi":"10.13026/f5mp-gx65","version_doi":"10.13026/n6yq-fq90","is_latest_version":true,"short_description":"Holter ECG database from Japan, containing data from 100 unique patients with paroxysmal AF including expert annotations of Supraventricular arrhythmias at the beat level.","publish_date":"2025-04-16","title":"SHDB-AF: a Japanese Holter ECG database of atrial fibrillation","abstract":"Saitama Heart Database Atrial Fibrillation (SHDB-AF) is a novel open-sourced Holter ECG database from Japan, containing data from 122 unique subjects with paroxysmal atrial fibrillation. Among the 128 recordings, 98 contain raw ECG data with rhythm annotations at the beat level, manually performed by a cardiology fellow. The remaining recordings consist only of ECG traces without annotations. The dataset was collected as part of a study evaluating the generalization performance of a deep learning atrial fibrillation event detection model across different distribution shifts.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":8887271676,"compressed_storage_size":7816783253,"resource_type":"Database","access_policy":"Open","topics":["atrial fibrillation","ecg","holters"],"source_url":"https://physionet.org/content/shdb-af/1.0.1/"},{"public_project_uuid":"ed7b8f50-7d46-4967-8697-06761d00c8f8","slug":"b2ai-voice","version":"2.0.0","core_doi":"10.13026/37yb-1t42","version_doi":"10.13026/3xt6-rf05","is_latest_version":false,"short_description":"A dataset of features from voice recordings and metadata to enable the development, benchmarking, and validation of clinically applicable machine-learning models for diagnosing a wide range of health conditions.","publish_date":"2025-04-16","title":"Bridge2AI-Voice: An ethically-sourced, diverse voice dataset linked to health information","abstract":"The human voice contains complex acoustic markers which have been linked to important health conditions including dementia, mood disorders, and cancer. When viewed as a biomarker, voice is a promising characteristic to measure as it is simple to collect, cost-effective, and has broad clinical utility. Recent advances in artificial intelligence have provided techniques to extract previously unknown prognostically useful information from dense data elements such as images. The Bridge2AI-Voice project seeks to create an ethically sourced flagship dataset to enable future research in artificial intelligence and support critical insights into the use of voice as a biomarker of health. Here we present Bridge2AI-Voice, a comprehensive collection of data derived from voice recordings with corresponding clinical information.
\nBridge2AI-Voice v2.0 contains data for 19,271 recordings collected from 442 participants across five sites in North America. Participants were selected based on known conditions which manifest within the voice waveform including voice disorders, neurological disorders, mood disorders, and respiratory disorders. The release contains data considered low risk, including derivations such as spectrograms but not the original voice recordings. Detailed demographic, clinical, and validated questionnaire data are also made available.
\nAudio recordings are included on a companion release on PhysioNet with the title \"Bridge2AI-Voice: An ethically-sourced, diverse voice dataset linked to health information (Audio Included)\". Please see that project for details to request access.
","license":{"name":"Bridge2AI Voice Registered Access License"},"dua":{"name":"Bridge2AI Voice Registered Access Agreement"},"main_storage_size":12438038640,"compressed_storage_size":12411200247,"resource_type":"Database","access_policy":"Credentialed","topics":["bridge2ai","voice"],"source_url":"https://physionet.org/content/b2ai-voice/2.0.0/"},{"public_project_uuid":"6ce6eb07-cbcd-4b99-978b-ccb15283315f","slug":"mimic-ext-synth-trial-question","version":"1.0.0","core_doi":"10.13026/rqxd-sz93","version_doi":"10.13026/30k0-av04","is_latest_version":true,"short_description":"In our recent study, we used Llama-3.1-70B-Instruct to generate synthetic training examples resembling clinical trial eligibility criteria. We manually reviewed 1000 of these examples and release them here.","publish_date":"2025-04-22","title":"MIMIC-III-Ext-Synthetic-Clinical-Trial-Questions","abstract":"Large-language models (LLMs) show promise for extracting information from clinical notes. Deploying these models at scale can be challenging due to high computational costs, regulatory constraints, and privacy concerns. To address these challenges, synthetic data distillation can be used to fine-tune smaller, open-source LLMs that achieve performance similar to the teacher model. These smaller models can be run on less expensive local hardware or at a vastly reduced cost in cloud deployments.
\nIn our recent study [1], we used Llama-3.1-70B-Instruct to generate synthetic training examples in the form of question-answer pairs along with supporting information. We manually reviewed 1000 of these examples and release them here. These examples can then be used to fine-tune smaller versions of Llama to improve their ability to extract clinical information from notes.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":12987862,"compressed_storage_size":4580777,"resource_type":"Database","access_policy":"Restricted","topics":["large language models","synthetic data distillation","clinical trial eligibility"],"source_url":"https://physionet.org/content/mimic-ext-synth-trial-question/1.0.0/"},{"public_project_uuid":"a326b5ea-2cf0-4786-8682-18f619c73704","slug":"swiss-mammo","version":"1.0.0","core_doi":"10.13026/gqzb-fg27","version_doi":"10.13026/mrg5-ja22","is_latest_version":false,"short_description":"Swiss-Mammo: A physician-written, synthetic dataset of 28 German mammography reports. The dataset is stratified based on BI-RADS categories and available in German and English.","publish_date":"2025-04-30","title":"Swiss-Mammo: A physician-written, synthetic dataset of German mammography reports","abstract":"This dataset, Swiss-Mammo, contains 28 manually constructed German mammography reports, each paired with an English translation. The reports are stratified across BI-RADS categories 0 through 6, with three reports per category. All reports were manually created by a radiology resident without the use of generative AI and independently reviewed by a senior radiologist specialized in breast imaging to ensure clinical plausibility and linguistic accuracy.
\nSwiss-Mammo was developed to support research on information extraction from radiology reports using large language models (LLMs) and frame semantics. The dataset provides a high-quality, controlled benchmark for evaluating systems designed to structure free-text radiological narratives. It is particularly suited for applications in natural language processing, clinical informatics, and decision support systems in breast imaging.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":98458,"compressed_storage_size":58008,"resource_type":"Database","access_policy":"Restricted","topics":["mammography","radiology","structured reporting","bi-rads"],"source_url":"https://physionet.org/content/swiss-mammo/1.0.0/"},{"public_project_uuid":"9d4dbb4b-2416-4834-a488-0cfb79b1c565","slug":"ann-pt-summ","version":"1.0.1","core_doi":"10.13026/m6hf-dq94","version_doi":"10.13026/gedc-j464","is_latest_version":true,"short_description":"Annotations for unsupported facts in 100 original MIMIC patient summaries (discharge instructions) and hallucinations in 100 Large Language Model (LLM) generated patient summaries labeled by two medical experts.","publish_date":"2025-04-30","title":"Medical Expert Annotations of Unsupported Facts in Doctor-Written and LLM-Generated Patient Summaries","abstract":"Large language models in healthcare can generate informative patient summaries while reducing the documentation workload of healthcare professionals. However, these models are prone to producing hallucinations, that is, generating unsupported information, which is problematic in the sensitive healthcare domain. To better characterize unsupported facts in medical texts, we developed a rigorous labeling protocol. Following this protocol, two medical experts annotated unsupported facts in 100 doctor-written summaries from the MIMIC-IV-Note Discharge Instructions and hallucinations 100 LLM-generated patient summaries. Here, we are releasing two datasets based on these annotations: Hallucinations-MIMIC-DI and Hallucinations-Generated-DI. We find that using these datasets to train on hallucination-free examples effectively reduces hallucinations for both Llama 2 (2.60 to 1.55 hallucinations per summary) and GPT-4 (0.70 to 0.40). Furthermore, we created a preprocessed version of the MIMIC-IV-Notes Discharge Instructions, releasing both a full-context version (MIMIC-IV-Note-Ext-DI) and a version that only uses the Brief Hospital Course for context (MIMIC-IV-Note-Ext-DI-BHC).
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":3292403761,"compressed_storage_size":1131370378,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/ann-pt-summ/1.0.1/"},{"public_project_uuid":"e620ff95-639a-404b-bc9d-857b5cf509be","slug":"mimic-iv-ext-apixaban-trial","version":"1.0.0","core_doi":"10.13026/k40z-bm16","version_doi":"10.13026/4p6q-vb04","is_latest_version":true,"short_description":"We created 23 questions resembling eligibility criteria from the apixaban clinical trial and evaluated them on a random sample of 100 patient notes from MIMIC-IV. We release the 2300 total question-answer pairs as a dataset here.","publish_date":"2025-04-30","title":"MIMIC-IV-Ext-Apixaban-Trial-Criteria-Questions","abstract":"Large-language models (LLMs) show promise for extracting information from clinical notes. Deploying these models at scale can be challenging due to high computational costs, regulatory constraints, and privacy concerns. To address these challenges, synthetic data distillation can be used to fine-tune smaller, open-source LLMs that achieve performance similar to the teacher model. These smaller models can be run on less expensive local hardware or at a vastly reduced cost in cloud deployments. In our recent study, we used Llama-3.1-70B-Instruct to generate synthetic training examples in the form of question-answer pairs along with supporting information. We then used these questions to fine-tune smaller versions of Llama to improve their ability to extract clinical information from notes.
\nTo evaluate the resulting models, we created 23 questions resembling eligibility criteria from the apixaban clinical trial and evaluated them on a random sample of 100 patient notes from MIMIC-IV. Notes from MIMIC-IV were taken from after 2012 to ensure no overlap with any of the notes from MIMIC-III which were used to generate the data used to finetune the models. We release the 2300 total question-answer pairs as a dataset here.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":33192103,"compressed_storage_size":914981,"resource_type":"Database","access_policy":"Restricted","topics":["clinical q and a evaluation set","clinical trial eligibility"],"source_url":"https://physionet.org/content/mimic-iv-ext-apixaban-trial/1.0.0/"},{"public_project_uuid":"5eb903ff-e2a7-48b9-8bda-da42e098e0b3","slug":"dreamt","version":"2.1.0","core_doi":"10.13026/dztc-dv77","version_doi":"10.13026/7r9r-7r24","is_latest_version":false,"short_description":"We present high resolution wearable device multichannel data along with clinical labeled and recorded sleep stage and polysomnography (PSG) data from 100 sleep abnormal patients with sleep apnea.","publish_date":"2025-04-30","title":"DREAMT: Dataset for Real-time sleep stage EstimAtion using Multisensor wearable Technology","abstract":"Sleep is an intrinsic part of human life, and recent advancements in wearable technology and machine learning have promised continuous and non-invasive methods of tracking sleep health and patterns, providing an important facet to a more holistic understanding of well-being. However, it is still challenging to achieve consistent and reliable real-time estimates of sleep stages using only smartwatches. This is especially true for individuals with irregular sleep patterns or sleep disorders. A major contributing factor is the distinct lack of publicly accessible, large-scale datasets that allow researchers and engineers to validate their wearable sleep staging algorithms against a population with diverse sleep patterns. Here, we present DREAMT, Dataset for Real-time sleep stage EstimAtion using Multisensor wearable Technology, a new dataset collected from 100 participants, which includes high-resolution signals from a smartwatch, expert sleep technician-annotated sleep stage labels, and clinical metadata related to sleep health and disorders.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":159249464551,"compressed_storage_size":25564322145,"resource_type":"Database","access_policy":"Restricted","topics":["sleep disorders","wearable","biomedical","time series classification"],"source_url":"https://physionet.org/content/dreamt/2.1.0/"},{"public_project_uuid":"72a78aef-eaff-4643-bf95-4b04d0903df0","slug":"minute-level-step-count-nhanes","version":"1.0.1","core_doi":"10.13026/5htm-7515","version_doi":"10.13026/9n0r-tv02","is_latest_version":false,"short_description":"Minute level step counts obtained from five step counting algorithms for raw accelerometry data, and minute level Activity Counts, MIMS, wear predictions, and wear flags for all participants who wore accelerometers in NHANES 2011-2014.","publish_date":"2025-05-05","title":"Minute level step counts and physical activity data from the National Health and Nutrition Examination Survey (NHANES) 2011-2014","abstract":"The National Health and Nutrition Examination Survey (NHANES) is a nationally representative study that collects demographic, socioeconomic, dietary, and health-related information from 10,000 Americans annually. In Wave G (2011-2012) and Wave H (2013-2014), participants wore an ActiGraph GT3X+ on the non-dominant wrist for seven consecutive days. Publicly available data include minute-level wear predictions (wake wear, sleep wear, unknown, nonwear), data quality flags, and monitor-independent movement summary units (MIMS). Additionally, raw triaxial accelerometer data recorded at 80 Hz are provided, enabling detailed analyses beyond MIMS.
\nWe applied five step counting algorithms and the ActiGraph Activity Count (AC) algorithm to the raw data to create a dataset of minute-level step counts. The dataset also includes minute level AC, MIMS, wear predictions, and wear flags for all participants who wore accelerometers in NHANES 2011–2014. These data facilitate detailed investigations of accelerometry-measured physical activity patterns in a nationally representative U.S. population. Potential analyses include examination of within-day activity patterns and their association with health outcomes.
","license":{"name":"Creative Commons Zero 1.0 Universal Public Domain Dedication"},"dua":null,"main_storage_size":2865163804,"compressed_storage_size":2865040605,"resource_type":"Database","access_policy":"Open","topics":["steps","accelerometry","physical activity","nhanes"],"source_url":"https://physionet.org/content/minute-level-step-count-nhanes/1.0.1/"},{"public_project_uuid":"877b5c47-92fc-4ec4-b439-197e5cb3d2b6","slug":"medisumqa","version":"1.0.0","core_doi":"10.13026/tvmt-8d97","version_doi":"10.13026/f566-h049","is_latest_version":true,"short_description":"MeDiSumQA is a dataset of patient-oriented QA pairs from MIMIC-IV discharge summaries, designed to evaluate LLMs in generating safe, patient-friendly medical responses for clinical QA and healthcare communication.","publish_date":"2025-05-05","title":"MeDiSumQA: Patient-Oriented Question-Answer Generation from Discharge Letters","abstract":"While increasing patients' access to medical documents improves medical care, this benefit is limited by varying health literacy levels and complex medical terminology. Large language models (LLMs) offer solutions by simplifying medical information. However, evaluating LLMs for safe and patient-friendly text generation is difficult due to the lack of standardized evaluation resources. To fill this gap, we developed MeDiSumQA. MeDiSumQA is a dataset created from MIMIC-IV discharge summaries through an automated pipeline combining LLM-based question-answer generation with manual quality checks. We use this dataset to evaluate various LLMs on patient-oriented question-answering. Our findings reveal that general-purpose LLMs frequently surpass biomedical-adapted models, while automated metrics correlate with human judgment. By releasing MeDiSumQA on PhysioNet, we aim to advance the development of LLMs to enhance patient understanding and ultimately improve care outcomes.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":4770256,"compressed_storage_size":1650692,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/medisumqa/1.0.0/"},{"public_project_uuid":"428c9fc5-0caa-44a7-a539-9be14d66a9a5","slug":"fdtooth","version":"1.0.0","core_doi":"10.13026/0q5t-zw54","version_doi":"10.13026/v9xk-dy61","is_latest_version":true,"short_description":"FDTooth is a dataset containing intraoral photographs and cone-beam computed tomography (CBCT) images with annotations for automated detection of fenestration and dehiscence in anterior teeth.","publish_date":"2025-05-05","title":"FDTooth: Intraoral Photographs and Cone-Beam Computed Tomography Images for Fenestration and Dehiscence Detection","abstract":"FDTooth is a comprehensive dataset designed for the automated detection of fenestration and dehiscence (FD) in anterior teeth, combining intraoral photographs and corresponding cone-beam computed tomography (CBCT) images from 241 patients aged 9 to 55 years. The dataset includes 1,800 annotated bounding boxes for intraoral photographs, with cone-beam computed tomography (CBCT) serving as the imaging modality of choice for assessing bone defects. It supports interdisciplinary dental research by enabling the development of deep learning models for non-invasive FD screening, providing an alternative assessment method that reduces radiation exposure. FDTooth serves as a valuable resource for advancing dental diagnostics, training professionals, and exploring other applications in dental and maxillofacial imaging.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":44160727911,"compressed_storage_size":44150049005,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/fdtooth/1.0.0/"},{"public_project_uuid":"3bf26cae-7509-469d-918e-ced244c0cc70","slug":"mimic-iv-ext-cardiac-disease","version":"1.0.0","core_doi":"10.13026/6dnk-7r25","version_doi":"10.13026/khgm-hc33","is_latest_version":true,"short_description":"The subset of the MIMIC-IV dataset includes the examination results and diagnostic information of 4,761 cardiac disease patients. The examination results for each patient are listed separately as evidence for the final diagnosis.","publish_date":"2025-05-06","title":"MIMIC-IV-Ext Cardiac Disease","abstract":"With the rapid development of generative LLMs (large language models) in the field of natural language processing, their potential in medical applications has become increasingly evident. However, most existing studies rely on exam-style questions or artificially designed cases, lacking validation using real patient data. To address this gap, this study leverages the MIMIC-IV database to construct a subset, MIMIC-IV-Ext Cardiac Disease, which includes 4,761 patients diagnosed with cardiac diseases. The dataset covers all relevant clinical examinations from admission to discharge, as well as the final diagnoses.Combining these data with the multi-turn interaction framework we built can be used to test whether large models can guide patients through in-hospital examinations. Moreover, after modifying the MIMIC-IV dataset, our sub-dataset can greatly facilitate researchers in conducting other studies.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":211010231,"compressed_storage_size":27198085,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimic-iv-ext-cardiac-disease/1.0.0/"},{"public_project_uuid":"6cce1635-d22a-493c-89f1-4306cf38984f","slug":"bigp3bci","version":"1.0.0","core_doi":"10.13026/ckte-3g83","version_doi":"10.13026/0byy-ry86","is_latest_version":true,"short_description":"A collection of data from P300-based brain-computer interface studies.","publish_date":"2025-05-19","title":"bigP3BCI: An Open, Diverse and Machine Learning Ready P300-based Brain-Computer Interface Dataset","abstract":"Brain–computer interfaces (BCIs) have wide-ranging applications as solutions for replacing or substituting neural output that has been lost because of severe neuromuscular injury or disease, such as individuals with late-stage amyotrophic lateral sclerosis (ALS). The P300-based BCI is one of the most commonly researched BCI for communication. This BCI dataset is curated from data originally generated from previous visual P300-based BCI speller studies, which include single- and multi-session experiments under a wide range of conditions. The BCI data are provided in an enriched and standardised format with BCI data elements that align with developing IEEE P2731 Working Group standards for BCI data to facilitate reusability. The data files, provided in open European Data Format ‘plus’, contain: i) electroencephalography (EEG) signals; ii) the BCI encoder, target characters and stimulus event markers for P300 event related potential analysis; iii) BCI spelling outcomes and feedback event markers for error related potential analysis; and if available, iv) self-reported demographics (age, sex, race, ethnicity); v) ALS diagnosis and a revised ALS Functional Rating Scale score obtained from medical records; and vi) eye tracker signals.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":47887839434,"compressed_storage_size":11077364408,"resource_type":"Database","access_policy":"Open","topics":["ieee p2731 working group standard","amyotrophic lateral sclerosis","p300 speller","p300 event related potential","oddball paradigm","error-related potential","electroencephalography","brain-computer interface"],"source_url":"https://physionet.org/content/bigp3bci/1.0.0/"},{"public_project_uuid":"29b463a8-ac22-48a8-b080-8dfcbcf09efd","slug":"hillel-yaffe-glaucoma-dataset","version":"1.0.0","core_doi":"10.13026/pdxv-m215","version_doi":"10.13026/z0ak-km33","is_latest_version":false,"short_description":"HYGD is a rigorously annotated fundus image dataset with gold-standard clinical labels designed to improve and benchmark deep learning models for accurate glaucoma detection.","publish_date":"2025-06-03","title":"Hillel Yaffe Glaucoma Dataset (HYGD): A Gold-Standard Annotated Fundus Dataset for Glaucoma Detection","abstract":"Glaucomatous optic neuropathy (GON) is a leading cause of irreversible blindness worldwide, affecting an estimated 64.3 million people globally with projections reaching 111.8 million by 2040. Approximately 50% of cases remain undiagnosed until advanced stages when vision loss becomes noticeable. Traditional diagnosis requires comprehensive ophthalmic examinations by specialists, creating accessibility barriers in many regions.
\nThe Hillel Yaffe Glaucoma Dataset (HYGD) addresses a critical limitation in existing GON datasets: the lack of gold-standard annotations. Unlike most publicly available datasets where glaucoma labels are determined solely from digital fundus images (DFIs), HYGD's labels are based on comprehensive ophthalmic examinations, including visual acuity assessment, intraocular pressure measurement, optical coherence tomography (OCT), visual field tests, and at least one year of follow-up monitoring.
\nThe dataset is structured to include both the DFIs and a labels file containing patient IDs, GON classifications and image quality scores. All DFIs were taken using a TOPCON DRI OCT Triton retinal camera with a 45° FOV and underwent deidentification and standardization processing.
\nHYGD enables researchers to train and benchmark models on rigorously annotated data, potentially improving their reliability. This dataset serves as a valuable resource for developing generalizable models that can function across diverse patient populations and clinical settings, ultimately supporting earlier detection and treatment of GON.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":131865715,"compressed_storage_size":124214721,"resource_type":"Database","access_policy":"Open","topics":["gon","ophthalmology","dfi","gold-standard","fundus","retina","glaucoma"],"source_url":"https://physionet.org/content/hillel-yaffe-glaucoma-dataset/1.0.0/"},{"public_project_uuid":"6e8961fe-e272-4304-8f3a-52d7d4445209","slug":"nacef","version":"2.0.0","core_doi":"10.13026/14tj-m189","version_doi":"10.13026/84f3-sa05","is_latest_version":false,"short_description":"Community-Acquired Pneumonia (CAP) poses a significant health risk, linked to high in-hospital morbidity and mortality rates. The dataset includes clinical details of 768 CAP patients at Clinica Universidad de La Sabana, Colombia.","publish_date":"2025-06-09","title":"Community-Acquired Pneumonia, Endotypes and Phenotypes (NACef): Prospective, observational cohort study of Translational Medicine","abstract":"Community-Acquired Pneumonia (CAP) remains a prominent infectious process associated with elevated in-hospital morbidity and mortality rates. Through the exploration of phenotypes, endotypes, and biomarkers, it becomes feasible to identify individuals at a heightened risk of adverse outcomes from CAP. The dataset encompasses clinical information from 768 patients diagnosed with CAP at Clinica Universidad de La Sabana, Colombia. Clinical data encompasses Baseline Clinical Data, In-hospital follow-up and data after hospital discharge. This repository has been utilized within the framework of a prospective research study, involving an observational cohort conducted in the domain of translational medicine. The dataset offers the opportunity for diverse statistical analyses, contributing to an enhanced comprehension of CAP, Endotypes, and Phenotypes. Moreover, this dataset proves valuable in the context of educational initiatives.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":208064,"compressed_storage_size":38553,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/nacef/2.0.0/"},{"public_project_uuid":"0a4f4537-5361-4f22-a459-20f7b2cd48bc","slug":"micro-nav","version":"1.1.0","core_doi":"10.13026/tyj5-dd44","version_doi":"10.13026/aeqv-0v88","is_latest_version":false,"short_description":"This study explores the respiratory microbiome's role in nosocomial lower respiratory tract infections in ICU patients. Conducted in Chía, Colombia, revealing the microbiome's impact on disease progression.","publish_date":"2025-06-24","title":"Microbiological, Immunological and Biochemical Characteristics of the Development of Ventilator Associated Pneumonia","abstract":"The respiratory microbiome plays a critical role in metabolism, immune system maturation, and protection against pathogens. Traditionally, respiratory microbiology in pneumonia focused on identifying a specific pathogen, often disregarding normal oral flora as contaminants. However, recent research highlights the importance of the pulmonary microbiome's altered composition (dysbiosis) and its association with diseases such as Ventilator-Associated Pneumonia (VAP). The relationship between dysbiosis and immune dysregulation remains unclear, raising questions about whether dysbiosis is a cause or consequence of disease progression. This dataset explores the respiratory microbiome's role in nosocomial lower respiratory tract infections in ICU patients.
\nThis prospective, observational dataset was built in an ICU in Chia, Colombia, focusing on adult patients requiring mechanical ventilation without initial pneumonia. Clinical data and biological samples, including bronchoalveolar lavage, orotracheal secretion, and blood, were collected at multiple time points. Microbiome analysis involved DNA extraction, gene sequencing of the 16S ribosomal unit, and bioinformatic analysis to characterize microbial diversity. Concurrently, pulmonary, and systemic inflammatory markers were quantified using ELISA.
\nThe dataset enrolled 141 patients, generating a comprehensive dataset on demographic, clinical, and microbiome variables. This research not only provides insights into the microbial ecology of VAP but also establishes a repository of data for future studies. The findings underscore the complex interplay between the microbiome and disease in critically ill patients, with implications for improving clinical outcomes in the ICU setting.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":5075530,"compressed_storage_size":396589,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/micro-nav/1.1.0/"},{"public_project_uuid":"755848f6-8822-4ba5-8921-753227c1a44e","slug":"swiss-mammo","version":"1.0.1","core_doi":"10.13026/gqzb-fg27","version_doi":"10.13026/95cy-rc09","is_latest_version":true,"short_description":"Swiss-Mammo: A physician-written, synthetic dataset of 28 German mammography reports. The dataset is stratified based on BI-RADS categories and available in German and English.","publish_date":"2025-06-24","title":"Swiss-Mammo: A physician-written, synthetic dataset of German mammography reports","abstract":"This dataset, Swiss-Mammo, contains 28 manually constructed German mammography reports, each paired with an English translation. The reports are stratified across BI-RADS categories 0 through 6, with four reports per category. All reports were manually created by a radiology resident without the use of generative AI and independently reviewed by a senior radiologist specialized in breast imaging to ensure clinical plausibility and linguistic accuracy.
\nSwiss-Mammo was developed to support research on information extraction from radiology reports using large language models (LLMs) and frame semantics. The dataset provides a high-quality, controlled benchmark for evaluating systems designed to structure free-text radiological narratives. It is particularly suited for applications in natural language processing, clinical informatics, and decision support systems in breast imaging.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":98458,"compressed_storage_size":58008,"resource_type":"Database","access_policy":"Restricted","topics":["mammography","structured reporting","radiology","bi-rads"],"source_url":"https://physionet.org/content/swiss-mammo/1.0.1/"},{"public_project_uuid":"74781ba9-13c9-4b12-a865-a09bf2445df5","slug":"tollet","version":"1.0.0","core_doi":"10.13026/z36n-vk32","version_doi":"10.13026/v66k-sk82","is_latest_version":false,"short_description":"We present tOLIet, the first thigh ECG dataset with real signals captured by a toilet seat with electrodes. There are 149 recordings from 86 people, useful for research into cardiovascular assessment using \"invisible\" ECG.","publish_date":"2025-06-24","title":"tOLIet: Single-lead Thigh-based Electrocardiography Using Polimeric Dry Electrodes","abstract":"Our team previously introduced an innovative concept for an \"invisible\" Electrocardiography (ECG) system, incorporating electrodes and sensors into a toilet seat design to enable signal acquisition from the thighs. Building upon that work, we now present a novel dataset featuring real-world, single-lead ECG signals captured at the thighs, offering a valuable resource for advancing research on thigh-based ECG for cardiovascular disease assessment. To our knowledge, this is the first dataset of its kind.
\nThe tOLIet dataset comprises 149 ECG recordings collected from 86 individuals (50 females, 36 males) with an average age of 31.73 ± 13.11 years, a mean weight of 66.89 ± 10.70 kg, and an average height of 166.82 ± 6.07 cm. Participants were recruited through direct contact with the Principal Investigator at Centro Hospitalar Universitário de Lisboa Central (CHULC) and via clinical consultations conducted at the same institution. Each recording includes four differential signals acquired from electrode pairs embedded in the toilet seat, with reference signals obtained from a standard 12-lead hospital ECG system.
","license":{"name":"Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":545345425,"compressed_storage_size":78228055,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/tollet/1.0.0/"},{"public_project_uuid":"bebd0b61-cb2c-4a49-910d-4742f2b9e70a","slug":"wearable-device-dataset","version":"1.0.1","core_doi":"10.13026/67z9-jh94","version_doi":"10.13026/he0v-tf17","is_latest_version":true,"short_description":"Physiological signals(Electrodermal Activity,Blood Volume Pulse, Heart Rate, Temperature,etc) from 36 healthy volunteers collected during structured acute stress induction and aerobic/anaerobic exercise sessions using the Empatica E4 wearable device.","publish_date":"2025-06-24","title":"Wearable Device Dataset from Induced Stress and Structured Exercise Sessions","abstract":"This original dataset contains physiological signals collected during structured acute stress induction and aerobic and anaerobic exercise sessions using a wearable device. Blood volume pulse, motion-based activity, skin temperature, and electrodermal activity were recorded with the Empatica E4, a research-grade wearable. The stress induction protocol involved math and emotional tasks designed to provoke stress responses, interleaved with rest periods. Self-reported stress levels were also recorded during this procedure. For the exercise sessions, distinct routines on a stationary bike were created for aerobic and anaerobic activities. The dataset includes records from 36 healthy volunteers for stress sessions, 30 for aerobic exercise, and 31 for anaerobic exercise. By examining the variations in physiological signals, the effects of these activities can be analyzed. This dataset is a valuable resource for research on stress and exercise detection and classification.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":259375714,"compressed_storage_size":73121143,"resource_type":"Database","access_policy":"Open","topics":["exercise","wearable","aerobic","stress","anaerobic"],"source_url":"https://physionet.org/content/wearable-device-dataset/1.0.1/"},{"public_project_uuid":"4b604095-70e5-4bdb-bce8-49d288ae1ee2","slug":"snomed-ct-entity-challenge","version":"1.1.0","core_doi":"10.13026/3f69-tk39","version_doi":"10.13026/qn8t-6e19","is_latest_version":false,"short_description":"272 discharge notes from the MIMIC-IV-Note dataset annotated with SNOMED CT concepts.","publish_date":"2025-07-22","title":"SNOMED CT Entity Linking Challenge","abstract":"This challenge, sponsored by SNOMED International, seeks to advance the development of Entity Linking models that operate on unstructured clinical texts. Participants in the challenge will train entity linking models using a subset of MIMIC-IV-Note discharge summaries that have been annotated with SNOMED CT concepts by a team of medical professionals. The full dataset (which is comprised of a training set and a test set) consists of approximately 75,000 annotations across nearly 300 discharge summaries.
\nThe challenge was originally run as a competition on the DrivenData platform. Now that the competition has concluded, the hidden test set has been added to this dataset.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":5361855,"compressed_storage_size":1636604,"resource_type":"Challenge","access_policy":"Credentialed","topics":["entity linking","clinical annotation","snomed"],"source_url":"https://physionet.org/content/snomed-ct-entity-challenge/1.1.0/"},{"public_project_uuid":"bb39665e-b379-47f6-b294-fe0a5933514a","slug":"mimic-ext-cxr-qba","version":"1.0.0","core_doi":"10.13026/6193-he91","version_doi":"10.13026/8qmz-da41","is_latest_version":false,"short_description":"We present a large-scale CXR VQA dataset derived from MIMIC-CXR with 42M QA pairs, featuring hierarchical answers, bounding boxes, and structured tags. We generated QA-pairs using LLM-based extraction from radiology reports and localization models.","publish_date":"2025-07-23","title":"MIMIC-Ext-CXR-QBA: A Structured, Tagged, and Localized Visual Question Answering Dataset with Question-Box-Answer Triplets and Scene Graphs for Chest X-ray Images","abstract":"Visual Question Answering (VQA) enables flexible and context-dependent analysis of medical images, such as chest X-rays (CXRs), by allowing users to pose specific questions and receive nuanced answers. However, existing CXR VQA datasets are typically limited to short and simplistic answer, lack localization information (such as bounding boxes), and provide little structured metadata (e.g., hierarchical answer formats or tags like region and finding annotations). To address these limitations, we introduce MIMIC-Ext-CXR-QBA, a new large-scale CXR VQA dataset derived from MIMIC-CXR, comprising 42 million QA pairs, which provides multi-granular, hierarchical answers composed of full sentences in the style of radiology reports, as well as detailed bounding boxes, and structured tags. Additionally, we provide scene graphs for each study, containing both regions and observation nodes with bounding boxes, tags, and textual descriptions derived from the original radiology reports. We created the scene graphs using LLM-based information extraction, semantic mention mapping, and localization models before generating question-answer pairs based on the extracted information stored in these graphs. Using automatic quality assessments, we have selected 31,230,906 QA pairs intended for pre-training and 7,532,281 of these intended for fine-tuning VQA models, therefore providing, to the best of our knowledge, the most sophisticated and largest VQA dataset for CXRs yet.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":28017838528,"compressed_storage_size":27215038197,"resource_type":"Database","access_policy":"Credentialed","topics":["localization","chest x-rays","scene graphs","vqa"],"source_url":"https://physionet.org/content/mimic-ext-cxr-qba/1.0.0/"},{"public_project_uuid":"82594569-5ed8-4c98-98bb-8766e892504b","slug":"sdoh-adverse-pregnancy","version":"1.0.0","core_doi":"10.13026/eahw-b275","version_doi":"10.13026/qk2y-wx30","is_latest_version":true,"short_description":"This project provides a manually annotated dataset of social determinants of health—social support, occupation, and substance use—linked to pregnancy outcomes, extracted from MIMIC-III and MIMIC-IV discharge summary notes.","publish_date":"2025-08-05","title":"Annotated Social Determinants of Health Dataset for Adverse Pregnancy Outcomes","abstract":"This project presents an annotated dataset derived from MIMIC-III and MIMIC-IV discharge summaries, focusing on key Social Determinants of Health (SDoH) factors—social support, occupation, and substance use—and their association with adverse pregnancy outcomes. Leveraging a combination of manual annotation and advanced Natural Language Processing (NLP) techniques, we developed and validated multiple models (rule-based, Word2Vec, and Clinical BERT) to automate the extraction of these features. The resulting de-identified dataset, along with our code scripts for data preprocessing, model development, and validation, is made publicly available through this PhysioNet project.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1743551,"compressed_storage_size":608221,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/sdoh-adverse-pregnancy/1.0.0/"},{"public_project_uuid":"59d7b597-82c3-4b41-ba0d-838e7e3d54d7","slug":"script-carpediem-dataset","version":"1.8.0","core_doi":"10.13026/1qke-qx33","version_doi":"10.13026/w7kd-qj96","is_latest_version":false,"short_description":"SCRIPT seeks to delineate the host/pathogen interactions during pneumonia using multiomic analysis of bronchoalveolar lavage fluid joined with clinical data and physician adjudication.","publish_date":"2025-08-05","title":"SCRIPT CarpeDiem Dataset: demographics, outcomes, and per-day clinical parameters for critically ill patients with suspected pneumonia","abstract":"Traditional approaches to analyzing episodes of patient care in the ICU examine features on presentation and outcomes on discharge, collapsing the numerous events that happen during a patient’s stay and ignoring intercurrent ICU complications or interventions. With CarpeDiem, we aimed to examine clinical features on a per-day basis as done during the common practice of daily physician rounds. The Successful Clinical Response in Pneumonia Therapy (SCRIPT) CarpeDiem Dataset features 12,495 patient-ICU-days from 585 patients enrolled in the SCRIPT study between June 2018 and March 2022, all of whom were critically ill patients on mechanical ventilation suspected of having pneumonia who underwent a bronchoalveolar lavage as part of routine clinical care. Each patient has demographics, pneumonia episode information as adjudicated by a panel of critical care physicians, outcomes, and 44 clinical parameters for which data are present including vital signs, laboratory parameters, and mechanical support devices. Data have been deidentified per Health Insurance Portability and Accountability Act (HIPAA) Safe Harbor rules. This dataset combines expert clinician adjudication with per-day granular information to provide a unique tool to examine the clinical courses of critically ill patients with pneumonia. We use this dataset in a manuscript examining the contribution of ventilator-associated pneumonia to patient outcomes (see Usage section for details) and share it for others to work with.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":20544955,"compressed_storage_size":3817190,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/script-carpediem-dataset/1.8.0/"},{"public_project_uuid":"7614c2c1-289a-4dac-946e-679ff364c434","slug":"immunosuppressive-annotations","version":"1.0.0","core_doi":"10.13026/tfay-qs74","version_doi":"10.13026/etd0-dq69","is_latest_version":true,"short_description":"This database contains 200 MIMIC-III admission notes with adjudicated labels for histories of various immunosuppressive conditions and usage of various immunosuppressive medications.","publish_date":"2025-08-05","title":"Immunosuppressive Condition and Medication Annotations for Admission Notes in the MIMIC-III Database","abstract":"Immunosuppression due to underlying conditions or immunosuppressive medication use increases the risk of morbidity and mortality in the context of infectious disease. Identifying patients with immunosuppression is important for better studying and understanding the impact of immunosuppression on critical care outcomes. While structured data (e.g., diagnosis codes, medication orders) from the electronic health record (EHR) can help identify patients with immunosuppression, the reliability of structured data is limited as it can miss more nuanced information that is only present in unstructured data, such as patient notes.
\nWe introduce a dataset for phenotyping immunosuppression, defined as identification of a patient’s immune status, based on admission notes. Patient admission notes were extracted from the Medical Information Mart for Intensive Care III (MIMIC-III) dataset, which contains health-related data and clinical notes associated with patients who stayed in critical care units at Beth Israel Deaconess Medical Center between 2001 and 2012. These notes were manually annotated for the presence of several immunosuppressive conditions and immunosuppressive medications.
\nEach admission note was independently annotated by two human annotators, and discrepancies were reviewed by an attending critical care physician. Annotated conditions include solid organ transplant, stem cell transplant, HIV, acute leukemia, lymphoma, multiple myeloma, and immunoglobulin deficiency. Annotated medications include azathioprine, cyclosporine, cyclophosphamide, mycophenolate, rituximab, and tacrolimus. This dataset can be leveraged for medical and computer science research, especially as related to the application of natural language processing and large language models (LLMs) in medicine. It can also be used as a starting point for research related to immunosuppression in critically ill patients.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":32764,"compressed_storage_size":11318,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/immunosuppressive-annotations/1.0.0/"},{"public_project_uuid":"a45df779-9b13-4c0d-af35-5b5ee5dc550e","slug":"echonext","version":"1.0.0","core_doi":"10.13026/2j37-ba56","version_doi":"10.13026/r9pp-3y42","is_latest_version":false,"short_description":"EchoNext is a curated dataset of electrocardiograms (ECGs) paired with echocardiogram-confirmed structural heart disease labels, designed to support the development and validation of machine learning models.","publish_date":"2025-08-05","title":"EchoNext: A Dataset for Detecting Echocardiogram-Confirmed Structural Heart Disease from ECGs","abstract":"This dataset contains a de-identified collection of 100,000 12-lead electrocardiograms (ECGs) with paired structural heart disease (SHD) labels derived from echocardiography, collected at Columbia University Irving Medical Center. Each ECG is provided with raw waveform data sampled at 250 Hz across all 12 leads, along with accompanying demographic and ECG-specific tabular metadata, including age, sex, heart rate, PR interval, QRS duration, and corrected QT interval. Each ECG is annotated with a binary label indicating the presence or absence of structural heart disease based on echocardiographic findings, including reduced left ventricular ejection fraction, increased ventricular wall thickness, significant valvular disease, right ventricular dysfunction, pulmonary hypertension, or pericardial effusion.
\nThis dataset was developed as part of the creation of the Columbia Mini-Model, a lightweight deep learning model for SHD detection from ECGs. The dataset represents a simplified, focused subset of the larger EchoNext training population and was used to evaluate model performance in resource-constrained settings or smaller-scale deployment environments. It is being released to promote transparency and reproducibility, support further research in cardiovascular AI, and enable benchmarking of lightweight ECG-based screening models for structural heart disease.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":24020960125,"compressed_storage_size":6486304836,"resource_type":"Database","access_policy":"Restricted","topics":["ai model deployment","population health","ecg","transthoracic echocardiogram","machine learning","left ventricular dysfunction","electrocardiogram","structural heart disease","aortic stenosis","cardiovascular screening","digital health","clinical decision support","artificial intelligence","ai in healthcare","health equity","deep learning","valvular heart disease","heart failure"],"source_url":"https://physionet.org/content/echonext/1.0.0/"},{"public_project_uuid":"5a017094-0353-4f72-8d3f-92ec5255b6d6","slug":"b2ai-voice","version":"2.0.1","core_doi":"10.13026/37yb-1t42","version_doi":"10.13026/gzjs-0535","is_latest_version":false,"short_description":"A dataset of features from voice recordings and metadata to enable the development, benchmarking, and validation of clinically applicable machine-learning models for diagnosing a wide range of health conditions.","publish_date":"2025-08-18","title":"Bridge2AI-Voice: An ethically-sourced, diverse voice dataset linked to health information","abstract":"The human voice contains complex acoustic markers which have been linked to important health conditions including dementia, mood disorders, and cancer. When viewed as a biomarker, voice is a promising characteristic to measure as it is simple to collect, cost-effective, and has broad clinical utility. Recent advances in artificial intelligence have provided techniques to extract previously unknown prognostically useful information from dense data elements such as images. The Bridge2AI-Voice project seeks to create an ethically sourced flagship dataset to enable future research in artificial intelligence and support critical insights into the use of voice as a biomarker of health. Here we present Bridge2AI-Voice, a comprehensive collection of data derived from voice recordings with corresponding clinical information.
\nBridge2AI-Voice v2.0 contains data for 19,271 recordings collected from 442 participants across five sites in North America. Participants were selected based on known conditions which manifest within the voice waveform including voice disorders, neurological disorders, mood disorders, and respiratory disorders. The release contains data considered low risk, including derivations such as spectrograms but not the original voice recordings. Detailed demographic, clinical, and validated questionnaire data are also made available.
\nAudio recordings are included on a companion release on PhysioNet with the title \"Bridge2AI-Voice: An ethically-sourced, diverse voice dataset linked to health information (Audio Included)\". Please see that project for details to request access.
","license":{"name":"Bridge2AI Voice Registered Access License"},"dua":{"name":"Bridge2AI Voice Registered Access Agreement"},"main_storage_size":12438038640,"compressed_storage_size":12411200247,"resource_type":"Database","access_policy":"Credentialed","topics":["bridge2ai","voice"],"source_url":"https://physionet.org/content/b2ai-voice/2.0.1/"},{"public_project_uuid":"6dddd3a5-07b4-403e-a7fc-7419c3c2b462","slug":"cxr-align","version":"1.0.0","core_doi":"10.13026/vvdj-2j30","version_doi":"10.13026/7ebc-s018","is_latest_version":true,"short_description":"CXR-Align is a benchmark dataset created to evaluate vision-language models' capability to interpret negations in chest X-ray (CXR) reports, featuring systematically modified reports from MIMIC-CXR.","publish_date":"2025-08-21","title":"CXR-Align: A Benchmark for CXR-Report Alignment with Negations","abstract":"CXR-Align is a benchmark dataset designed to evaluate vision-language processing (VLP) models' ability to accurately interpret negations in chest X-ray (CXR) reports. Negations are prevalent in medical documentation and pose significant challenges for automated analysis, as misinterpretation can lead to critical diagnostic errors. Existing medical VLP systems often inadequately handle negated findings, motivating the creation of CXR-Align to specifically target and mitigate this limitation.
\nThe dataset comprises systematically modified and anonymized clinical reports derived from the MIMIC-III database. Each report in CXR-Align includes controlled alterations, particularly negations introduced to clinically relevant positive findings. Reports are standardized to exclude normal CXR cases, ensuring that each entry maintains direct diagnostic significance and maximizes the datase's utility for assessing nuanced language comprehension.
\nCXR-Align enables researchers to rigorously test the accuracy and robustness of VLP models, advancing the development of systems capable of reliably interpreting complex clinical language. The dataset's structured format and clear diagnostic focus make it an essential resource for researchers aiming to enhance model comprehension of medical negations, ultimately contributing to safer and more effective clinical decision-support tools.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":2739462,"compressed_storage_size":336163,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/cxr-align/1.0.0/"},{"public_project_uuid":"8bc6846d-46b1-4fa5-838d-3eae0fccc625","slug":"orchid","version":"2.1.0","core_doi":"10.13026/b1c0-3506","version_doi":"10.13026/637d-3e59","is_latest_version":false,"short_description":"Multi-center dataset on organ procurement in the United States","publish_date":"2025-08-21","title":"Organ Retrieval and Collection of Health Information for Donation (ORCHID)","abstract":"There are well-documented inefficiencies and inequities in the current system of deceased donor organ transplantation. While much prior research has focused on designing better allocation systems to distribute donated organs, more can be done to study organ procurement, the process by which organs are recovered from deceased donors. Improvements to organ procurement can greatly increase the supply of a scarce public resource, improving health outcomes for the more than 100,000 patients currently waiting for an organ transplant. Such changes will also impact health equity, as a disproportionate number of waitlisted patients come from marginalized communities. So far, research into organ procurement has been hamstrung by a lack of transparent data. To address this pressing issue, we have collected the first ever multi-center dataset on organ procurement, covering ten years of clinical, financial, and administrative information from six organ procurement organizations (OPOs). This novel dataset can be used to identify opportunities to improve current practice, build tools to help OPOs make better decisions, and recommend policy changes to increase procurement.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":1054232476,"compressed_storage_size":125079198,"resource_type":"Database","access_policy":"Restricted","topics":["organ procurement organizations","organ transplantation"],"source_url":"https://physionet.org/content/orchid/2.1.0/"},{"public_project_uuid":"aefad4a4-179f-4a2f-a00c-1c801f06ff37","slug":"nacef","version":"2.0.1","core_doi":"10.13026/14tj-m189","version_doi":"10.13026/4y3t-pq44","is_latest_version":true,"short_description":"Community-Acquired Pneumonia (CAP) poses a significant health risk, linked to high in-hospital morbidity and mortality rates. The dataset includes clinical details of 768 CAP patients at Clinica Universidad de La Sabana, Colombia.","publish_date":"2025-08-22","title":"Community-Acquired Pneumonia, Endotypes and Phenotypes (NACef): Prospective, observational cohort study of Translational Medicine","abstract":"Community-Acquired Pneumonia (CAP) remains a prominent infectious process associated with elevated in-hospital morbidity and mortality rates. Through the exploration of phenotypes, endotypes, and biomarkers, it becomes feasible to identify individuals at a heightened risk of adverse outcomes from CAP. The dataset encompasses clinical information from 768 patients diagnosed with CAP at Clinica Universidad de La Sabana, Colombia. Clinical data encompasses Baseline Clinical Data, In-hospital follow-up and data after hospital discharge. This repository has been utilized within the framework of a prospective research study, involving an observational cohort conducted in the domain of translational medicine. The dataset offers the opportunity for diverse statistical analyses, contributing to an enhanced comprehension of CAP, Endotypes, and Phenotypes. Moreover, this dataset proves valuable in the context of educational initiatives.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":209450,"compressed_storage_size":38495,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/nacef/2.0.1/"},{"public_project_uuid":"cda953fd-9c3d-4ef3-94ab-7ffbd23ed19b","slug":"mimic-iv-fhir-demo","version":"2.1.0","core_doi":"10.13026/vphg-y548","version_doi":null,"is_latest_version":true,"short_description":"The MIMIC-IV Clinical Database Demo on FHIR is a 100 patient subset of the MIMIC-IV v2.2 and MIMIC-IV-ED v2.2 clinical databases converted into the Fast Healthcare Interoperability Resources (FHIR) format.","publish_date":"2025-08-27","title":"MIMIC-IV Clinical Database Demo on FHIR","abstract":"Interoperability of healthcare data has become increasingly important given the increase in deployment of data driven algorithms in clinical settings. The Fast Healthcare Interoperability Resources (FHIR) standard has emerged as a promising mechanism to share healthcare data across vendors in real-time and batch settings. Real-world datasets available in FHIR would accelerate research and development of data-driven algorithms. Existing datasets in FHIR are primarily synthetic, and cover a limited number of resources. To address this gap, we have reformatted the Medical Information Mart for Intensive Care (MIMIC)-IV Clinical Database Demo into FHIR. The MIMIC clinical databases have received wide adoption and the constituent data are understood by the community. As much as possible, we adhered to the base resources with minimal extensions. Alongside the dataset, we publish openly available code allowing researchers to quickly build upon our work. Translating MIMIC-IV into FHIR provides a benchmark dataset for institutions to experiment with FHIR based tools, and we hope this resource supports adoption and use of FHIR.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":51885265,"compressed_storage_size":51859726,"resource_type":"Database","access_policy":"Open","topics":["electronic health records","fhir","mimic"],"source_url":"https://physionet.org/content/mimic-iv-fhir-demo/2.1.0/"},{"public_project_uuid":"dca3a1d3-9594-4930-b779-a4b4daa17f3b","slug":"vital-file-web-viewer","version":"1.0.0","core_doi":"10.13026/cgsq-t636","version_doi":"10.13026/3642-e205","is_latest_version":true,"short_description":"A standalone, browser-based viewer for .vital files that enables interactive visualization of high-resolution vital signs data without server-side components.","publish_date":"2025-09-01","title":"Vital File Web Viewer: A Browser-Based Tool for Viewing High-Resolution Vital Signs Data","abstract":"Vital File Web Viewer is an open-source, browser-based tool for visualizing high-resolution vital signs data stored in the .vital format, commonly used during surgery or intensive care monitoring. The viewer enables seamless access to time-series waveform data across operating systems without requiring installation or server-side processing. This tool facilitates rapid data exploration and quality assessment in clinical research environments, particularly when centralized software infrastructure is unavailable.
","license":{"name":"Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":139642,"compressed_storage_size":140297,"resource_type":"Software","access_policy":"Open","topics":["vital file","vitaldb"],"source_url":"https://physionet.org/content/vital-file-web-viewer/1.0.0/"},{"public_project_uuid":"ba071bd9-5f93-4658-829e-9ff243062267","slug":"hillel-yaffe-fundus-amd","version":"1.0.0","core_doi":"10.13026/f0dn-8q46","version_doi":"10.13026/ydf1-z238","is_latest_version":true,"short_description":"The HYAMD dataset comprises 1,560 high-resolution fundus images from 325 patients, aimed at validating machine learning models for age-related macular degeneration (AMD) diagnosis.","publish_date":"2025-09-09","title":"HYAMD High-Resolution Fundus Image Dataset for age related macular degeneration (AMD) Diagnosis","abstract":"The Hillel Yaffe Age Related Macular Degeneration (HYAMD) longitudinal dataset comprises of 1,560 Digital Fundus Images (DFIs) of 325 patients examined at the Hillel Yaffe Medical Center (Hadera, Israel, Helsinki approval number 0048-24-HYMC) providing follow-up data from 2021 to 2024. The AMD cohort includes 147 patients (73 males, 74 females) aged 54–94, while the control group comprises 190 diabetic retinopathy (DR) patients (111 males, 78 females) aged 24–92. DFIs in the AMD cohort either present with early, intermediate, or late-stage AMD, including neovascular AMD (NVAMD), visible geographic atrophy (GA), or presence of large drusen (LD). Diagnosis was based on a comprehensive clinical ophthalmic examination and supported by Optical Coherence Tomography (OCT) scans, including OCT angiography. Non-AMD DFIs were taken from patients examined in the Retina clinic for DR and without a concurrent diagnosis of AMD. Diagnosis of DR was based on clinical examination, macular OCT, and in relevant cases fluorescein angiography, and widefield imaging. This dataset was developed for the validation of machine learning models intended to the task of AMD identification from single DFI. The annotations are gold-standard, meaning the AMD labels were determined based on a complete ophthalmic evaluation rather than solely on DFIs. The images were taken using a DRI OCT Triton (Topcon) camera, with a 45° field of view (FOV) and a resolution of 1960 × 1934 pixels. HYAMD is also, to the best of our knowledge, the first open access retinal dataset from an Israeli sample.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":2635455483,"compressed_storage_size":2628223745,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/hillel-yaffe-fundus-amd/1.0.0/"},{"public_project_uuid":"01efce8f-4dae-42f4-8bc8-6dfbb78cf5ed","slug":"mimic-iv-ext-instr","version":"1.0.0","core_doi":"10.13026/5f7e-9s36","version_doi":"10.13026/e5bq-pr14","is_latest_version":true,"short_description":"This dataset contains 450K open-ended instruction-following examples generated using GPT-3.5 based on the MIMIC-IV EHR database.","publish_date":"2025-09-09","title":"MIMIC-IV-Ext-Instr: A Dataset of 450K+ EHR-Grounded Instruction-Following Examples","abstract":"Large language models (LLMs) have shown impressive capabilities in solving a wide range of tasks based on human instructions. However, developing a conversational AI assistant for electronic health record (EHR) data remains challenging due to the lack of large-scale instruction-following datasets. To address this, we present MIMIC-IV-Ext-Instr, a dataset containing over 450K open-ended, instruction-following examples generated using GPT-3.5 on a HIPAA-compliant platform. Derived from the MIMIC-IV EHR database, MIMIC-IV-Ext-Instr spans a wide range of topics and is specifically designed to support instruction-tuning of general-purpose LLMs for diverse clinical applications.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":123383432,"compressed_storage_size":25261967,"resource_type":"Database","access_policy":"Credentialed","topics":["medical question answering","large language models","instruction tuning"],"source_url":"https://physionet.org/content/mimic-iv-ext-instr/1.0.0/"},{"public_project_uuid":"6114eaeb-2c1e-4ecf-9254-d3b5f9df41e3","slug":"myocardial-perfusion-spect","version":"1.0.0","core_doi":"10.13026/md6s-re23","version_doi":"10.13026/ce2z-dw74","is_latest_version":true,"short_description":"This database provides a collection of myocardial perfusion scintigraphy images. The dataset encompasses a diversity of clinical cases, including various perfusion patterns and underlying cardiac conditions.","publish_date":"2025-09-10","title":"Myocardial perfusion scintigraphy image database","abstract":"This database provides a collection of myocardial perfusion scintigraphy images in DICOM format with all metadata and segmentations (masks) in NIfTI format. The images were obtained from patients undergoing scintigraphy examinations to investigate cardiac conditions such as ischemia and myocardial infarction. The dataset encompasses a diversity of clinical cases, including various perfusion patterns and underlying cardiac conditions. All images have been properly anonymized, and the age range of the patients is from 20 to 90 years. This database represents a valuable source of information for researchers and healthcare professionals interested in the analysis and diagnosis of cardiac diseases. Moreover, it serves as a foundation for the development and validation of image processing algorithms and artificial intelligence techniques applied to cardiovascular medicine. Available for free on the PhysioNet platform, its aim is to promote collaboration and advance research in nuclear cardiology and cardiovascular medicine, while ensuring the replicability of studies.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":52824634,"compressed_storage_size":22492715,"resource_type":"Database","access_policy":"Open","topics":["myocardial perfusion","systems modeling","myocardial perfusion scintigraphy","dicom","metadata","artificial intelligence","ventricular walls","coronary artery disease","convolutional neural networks","automated segmentation","clinical diagnosis","anonymization","nifti"],"source_url":"https://physionet.org/content/myocardial-perfusion-spect/1.0.0/"},{"public_project_uuid":"2441efe5-477d-4104-97c0-38f754eefc40","slug":"mcphases","version":"1.0.0","core_doi":"10.13026/5b8h-jx61","version_doi":"10.13026/zx6a-2c81","is_latest_version":true,"short_description":"This initial version of the PHASES dataset includes multimodal menstrual health data—hormone levels, wearable sensor metrics, and self-reported symptoms—collected across two study intervals from 42 young adults.","publish_date":"2025-09-10","title":"mcPHASES: A Dataset of Physiological, Hormonal, and Self-reported Events and Symptoms for Menstrual Health Tracking with Wearables","abstract":"Individuals who menstruate are frequently led to believe that there is a standard menstrual cycle, typically characterized as 28 days in length with predictable and uniform patterns. This framing often emphasizes cycle dates as the only relevant metric, overlooking the broader physiological and emotional fluctuations throughout the cycle driven by complex hormonal interactions. Consequently, when individuals encounter menstrual experiences that do not align with calendar-based metrics, they are often left without adequate frameworks for understanding their menstrual health, which can result in distress or delays in seeking care. Our work advocates for a new definition of menstrual health that encompasses a wider range of physiological signals in order to acknowledge its connection to overall wellbeing, establish realistic expectations for menstruators, and build better health management systems. However, historical stigmatization has led to a dearth of datasets suitable for pursuing these aims.
\nmcPHASES (menstrual cycle Physiological, Hormonal, and Self-Reported Events and Symptoms) is a comprehensive dataset consisting of multimodal physiological, hormonal, and self-reported measures collected to support holistic menstrual health research. Data from 42 Canadian young adult menstruators was collected across two 3-month periods. Participants wore Fitbit Sense smartwatches and Dexcom G6 continuous glucose monitors to measure physiological signals, and they used Mira Plus Starter Kits to track their hormone levels. Additionally, participants self-reported daily experiences like cramps, sleep quality, and stress levels. The dataset contains 23 structured tables organized by signal category so that researchers can examine relationships between physiological signals and hormonal fluctuations, analyze the impacts of lifestyle factors on the menstrual cycle, and develop better algorithms for menstrual cycle prediction. More broadly, mcPHASES supports research in women's health, digital health technologies, and personalized care by providing unprecedented multimodal data for building a more accurate understanding of menstrual health patterns.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":3676157365,"compressed_storage_size":367098676,"resource_type":"Database","access_policy":"Restricted","topics":["hormones","menstrual health","multimodal health","wearables","health sensor data","womens health"],"source_url":"https://physionet.org/content/mcphases/1.0.0/"},{"public_project_uuid":"1a2542b8-095e-4dbf-8fe2-658d2bb2f4e3","slug":"radgraph-xl","version":"1.0.0","core_doi":"10.13026/6tw7-rq96","version_doi":"10.13026/j8e7-pr22","is_latest_version":true,"short_description":"RadGraph-XL is a large, expert-annotated dataset of 2,300 radiology reports covering multiple modalities and anatomies. It enables accurate extraction of clinical entities and relations for downstream medical AI tasks.","publish_date":"2025-09-12","title":"RadGraph-XL: A Large-Scale Expert-Annotated Dataset for Entity and Relation Extraction from Radiology Reports","abstract":"Radiology reports are essential for clinical care but pose challenges for automated processing due to their unstructured nature. Existing datasets like RadGraph-1.0 focus narrowly on chest X-rays (CXR), limiting their applicability. We introduce RadGraph-XL, a large-scale, expert-annotated dataset of 2,300 radiology reports with over 410,000 labeled entities and relations, spanning four anatomy-modality pairs: chest computed tomography (CT), abdomen/pelvis CT, brain magnetic resonance imaging (MR), and CXR.
\nEach report is annotated by board-certified radiologists using a detailed schema that captures observations, anatomical references, and their relationships. A novel post-processing step identifies measurement-related entities, a clinically valuable category. Trained models using RadGraph-XL outperform prior methods and GPT-4, and generalize well to out-of-domain data such as deep vein thrombosis (DVT) ultrasound reports.
\nRadGraph-XL is released publicly with models and annotations to support applications in clinical natural language processing (NLP), medical imaging artificial intelligence, and foundation model evaluation, setting a new benchmark for structured information extraction in radiology.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":3089420,"compressed_storage_size":526091,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/radgraph-xl/1.0.0/"},{"public_project_uuid":"9ea57e58-641f-4f2e-88c2-bab7a9fe278c","slug":"echonext","version":"1.1.0","core_doi":"10.13026/2j37-ba56","version_doi":"10.13026/3ykd-bf14","is_latest_version":false,"short_description":"EchoNext is a curated dataset of electrocardiograms (ECGs) paired with echocardiogram-confirmed structural heart disease labels, designed to support the development and validation of machine learning models.","publish_date":"2025-09-16","title":"EchoNext: A Dataset for Detecting Echocardiogram-Confirmed Structural Heart Disease from ECGs","abstract":"This dataset contains a de-identified collection of 100,000 12-lead electrocardiograms (ECGs) with paired structural heart disease (SHD) labels derived from echocardiography, collected at Columbia University Irving Medical Center. Each ECG is provided with raw waveform data sampled at 250 Hz across all 12 leads, along with accompanying demographic and ECG-specific tabular metadata, including age, sex, heart rate, PR interval, QRS duration, and corrected QT interval. Each ECG is annotated with a binary label indicating the presence or absence of structural heart disease based on echocardiographic findings, including reduced left ventricular ejection fraction, increased ventricular wall thickness, significant valvular disease, right ventricular dysfunction, pulmonary hypertension, or pericardial effusion.
\nThis dataset was developed as part of the creation of the Columbia Mini-Model, a lightweight deep learning model for SHD detection from ECGs. The dataset represents a simplified, focused subset of the larger EchoNext training population and was used to evaluate model performance in resource-constrained settings or smaller-scale deployment environments. It is being released to promote transparency and reproducibility, support further research in cardiovascular AI, and enable benchmarking of lightweight ECG-based screening models for structural heart disease.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":24023103425,"compressed_storage_size":6486764061,"resource_type":"Database","access_policy":"Restricted","topics":["aortic stenosis","deep learning","health equity","cardiovascular screening","valvular heart disease","heart failure","digital health","ecg","machine learning","ai model deployment","left ventricular dysfunction","artificial intelligence","clinical decision support","ai in healthcare","population health","electrocardiogram","transthoracic echocardiogram","structural heart disease"],"source_url":"https://physionet.org/content/echonext/1.1.0/"},{"public_project_uuid":"557b7f8e-933f-4ee6-a269-bded4500c82e","slug":"radvlm-instruction-dataset","version":"1.0.0","core_doi":"10.13026/6wf1-7382","version_doi":"10.13026/et5g-h222","is_latest_version":true,"short_description":"This dataset is designed to construct RadVLM, a vision–language model for chest X-ray interpretation. It includes instruction data for tasks such as report generation, abnormality detection, and region grounding, and multitask conversation.","publish_date":"2025-09-25","title":"RadVLM Instruction Dataset","abstract":"We release the RadVLM instruction dataset, a large-scale resource used to train the RadVLM model on diverse radiology tasks. The dataset contains 1,115,021 image–instruction pairs spanning five task families: (i) report generation from frontal CXRs using filtered Findings/Impression text; (ii) abnormality classification for the standard 14 CheXpert labels; (iii) anatomy grounding; (iv) abnormality detection and grounding; and (v) phrase grounding from report sentences. To support interactive use, we include ~89k LLM-generated multi-turn, multi-task conversations (~3k with spatial grounding) derived from image-linked attributes (reports, labels, boxes). Creation involved curating datasets from public sources, excluding lateral views, removing prior-study references and other non-image context from reports, fusing multi-reader annotations, and harmonizing label and coordinate formats. The resource is intended for training CXR assistants across diverse radiology tasks and within a conversational format.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":470326840,"compressed_storage_size":88873766,"resource_type":"Database","access_policy":"Credentialed","topics":["vision-language models","medical ai","chest x-rays"],"source_url":"https://physionet.org/content/radvlm-instruction-dataset/1.0.0/"},{"public_project_uuid":"d96a3a51-590b-4870-9cb9-d09e1f39b240","slug":"mimic-ext-drug-detection","version":"1.0.0","core_doi":"10.13026/e8ew-ed16","version_doi":"10.13026/0kyx-r485","is_latest_version":true,"short_description":"This project offers a multilabel annotated dataset of clinical note sentences from MIMIC-III/IV for substance use detection. It supports NLP research for identifying various co-occurring drug use mentions in patient records.","publish_date":"2025-09-25","title":"MIMIC-Ext-DrugDetection","abstract":"This project shares a large, annotated drug detection dataset created from MIMIC-III/IV discharge summaries. The dataset was developed to address the challenge of identifying substance use behaviors in Electronic Health Records (EHRs), where critical details are often embedded in unstructured notes requiring contextual interpretation. The primary aim was to support future systemic substance use surveillance. The data consists of medical notes tokenized into sentences, annotated for eight substance use categories: heroin, cocaine, methamphetamine, illicit use of prescription opioids and benzodiazepines, cannabis, Injection Drug Use (IDU), and general drug use. The dataset was used to evaluate the performance of various large language models (LLMs) for detecting these substance use categories, demonstrating that LLMs, particularly a fine-tuned model, can significantly enhance detection accuracy and show promise for clinical decision support and research.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1320786,"compressed_storage_size":396499,"resource_type":"Database","access_policy":"Credentialed","topics":["prescription opioid misuse","cannabis","benzodiazepine misuse","ehr","injection drug use","heroin","methamphetamine","substance use","multi-label","cocaine","drug detection","polysubstance use","mimic-iv","mimic-iii","clinical notes"],"source_url":"https://physionet.org/content/mimic-ext-drug-detection/1.0.0/"},{"public_project_uuid":"cd174a71-3f42-49bc-ab2a-bd49018f9740","slug":"mc-med","version":"1.0.1","core_doi":"10.13026/xgx1-7x47","version_doi":"10.13026/wvyw-g663","is_latest_version":true,"short_description":"A multimodal dataset of deidentified clinical and physiological data from emergency department visits, supporting research on patient outcomes, care processes, and the effects of continuous monitoring during and after the COVID-19 pandemic.","publish_date":"2025-09-25","title":"Multimodal Clinical Monitoring in the Emergency Department (MC-MED)","abstract":"Emergency department (ED) patients often present with undiagnosed complaints, and can exhibit rapidly evolving physiology. Therefore, data from continuous physiologic monitoring, in addition to the electronic health record, is essential to understand the acute course of illness and responses to interventions. The complexity of ED care and the large amount of unstructured multimodal data it produces has limited the accessibility of detailed ED data for research. We release Multimodal Clinical Monitoring in the Emergency Department (MC-MED), a comprehensive, multimodal, and de-identified clinical and physiological dataset. MC-MED includes 118,385 adult ED visits to an academic medical center from 2020 to 2022. Data include continuously monitored vital signs, physiologic waveforms (electrocardiogram, photoplethysmogram, respiration), patient demographics, medical histories, orders, medication administrations, laboratory and imaging results, and visit outcomes. MC-MED is the first dataset to combine detailed physiologic monitoring with clinical events and outcomes for a large, diverse ED population.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":515384410382,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mc-med/1.0.1/"},{"public_project_uuid":"34683243-8fc2-4428-9b8d-4e676dedcc39","slug":"mimic-iv-ext-22mcts","version":"1.0.0","core_doi":"10.13026/dmdq-4w10","version_doi":"10.13026/dkj6-r828","is_latest_version":true,"short_description":"It is a time series clinical events dataset with concrete temporal information. The dataset consists of 22,588,586 clinical events and related timestamps from 267,284 discharge summaries of the MIMIC-IV-Note.","publish_date":"2025-09-29","title":"MIMIC-IV-Ext-22MCTS: A 22 Millions-Event Temporal Clinical Time-Series Dataset with Relative Timestamp","abstract":"Clinical risk prediction based on machine learning algorithms plays a vital role in modern healthcare. A crucial component in developing a reliable prediction model is a high-quality dataset with time series clinical events. In this work, we release such a dataset that consists of 22,588,586 clinical time series events, which we term MIMIC-IV-Ext-22MCTS. Our source data are discharge summaries selected from the well-known yet unstructured MIMIC-IV-Note. We then extract clinical events as short text spans from the discharge summaries, along with the timestamps of these events as temporal information by contextual retrieval and Llama-3.1-8B.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":875824224,"compressed_storage_size":139291331,"resource_type":"Database","access_policy":"Credentialed","topics":["clinical event annotation","mimic","time series","temporal annotation"],"source_url":"https://physionet.org/content/mimic-iv-ext-22mcts/1.0.0/"},{"public_project_uuid":"7f536133-54f0-4f82-a69f-c71afaf35ecf","slug":"mimic-iv-demo-meds","version":"0.0.1","core_doi":"10.13026/y9xz-1347","version_doi":"10.13026/t2y8-ea41","is_latest_version":true,"short_description":"MIMIC-IV Clinical Database Demo in MEDS (Medical Event Data Standard) format.","publish_date":"2025-09-30","title":"MIMIC-IV demo data in the Medical Event Data Standard (MEDS)","abstract":"This dataset is an automated ETL conversion of the MIMIC-IV Clinical Database Demo into the Medical Event Data Standard (MEDS). MEDS is a data schema for storing streams of medical events such as those sourced from Electronic Health Records or claims records. MEDS is intentionally a minimal standard, designed for maximum interoperability across datasets, existing tools, and model architectures. By providing a simple standardization layer between datasets and model-specific code, MEDS is intended to help make machine learning research for EHR data more reproducible, robust, computationally performant, and collaborative.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":5976560,"compressed_storage_size":4923223,"resource_type":"Database","access_policy":"Open","topics":["electronic health record","mimic","meds","machine learning","critical care","medical event data standard","ehr"],"source_url":"https://physionet.org/content/mimic-iv-demo-meds/0.0.1/"},{"public_project_uuid":"a29d9f61-f13e-4982-989f-a16c5ffd4388","slug":"orchid","version":"2.1.1","core_doi":"10.13026/b1c0-3506","version_doi":"10.13026/rfeq-j318","is_latest_version":true,"short_description":"Multi-center dataset on organ procurement in the United States","publish_date":"2025-09-30","title":"Organ Retrieval and Collection of Health Information for Donation (ORCHID)","abstract":"There are well-documented inefficiencies and inequities in the current system of deceased donor organ transplantation. While much prior research has focused on designing better allocation systems to distribute donated organs, more can be done to study organ procurement, the process by which organs are recovered from deceased donors. Improvements to organ procurement can greatly increase the supply of a scarce public resource, improving health outcomes for the more than 100,000 patients currently waiting for an organ transplant. Such changes will also impact health equity, as a disproportionate number of waitlisted patients come from marginalized communities. So far, research into organ procurement has been hamstrung by a lack of transparent data. To address this pressing issue, we have collected the first ever multi-center dataset on organ procurement, covering ten years of clinical, financial, and administrative information from six organ procurement organizations (OPOs). This novel dataset can be used to identify opportunities to improve current practice, build tools to help OPOs make better decisions, and recommend policy changes to increase procurement.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":715700569,"compressed_storage_size":81807404,"resource_type":"Database","access_policy":"Restricted","topics":["organ procurement organizations","organ transplantation"],"source_url":"https://physionet.org/content/orchid/2.1.1/"},{"public_project_uuid":"87d54adc-10a4-4ca6-9c85-c62a3dc5a2db","slug":"tn-mammo-breast-density","version":"1.0.0","core_doi":"10.13026/34kz-bk76","version_doi":"10.13026/1kx0-xc60","is_latest_version":true,"short_description":"We release the first version of TN-Mammo (June 2024), a mammogram dataset of 676 cases with breast density labels, providing high-quality data to support machine learning and early breast cancer detection.","publish_date":"2025-10-04","title":"TN-Mammo: A Multi-view Mammography Dataset for Breast Density Classification","abstract":"Breast cancer is one of the most common types of cancer among women, leading to a growing and essential need for early and precise detection. A variety of machine learning techniques have been demonstrating great promise in improving diagnostic accuracy integrated with digital mammography, which remains the gold standard screening technique in the detection of early breast cancer. Since the performance of machine learning models generally depends on the quality of training data, it is essential to build a high-quality mammogram dataset. Although there exists a number of public datasets, there are still significant limitations due to both their quality and quantity. We present an extensive Vietnamese mammogram dataset with breast density annotations in an effort to bridge the gap between availability and practical usability. Our mammogram dataset named TNMammo, consists of bilateral craniocaudal (CC) and mediolateral oblique (MLO) views for 676 subjects, each with paired left and right breast views. The breast density and orientation for each case have been independently assessed by two radiologists in a double-blind manner, ensuring consistency and reliability in the annotations. Each case is evaluated based on breast density, categorized into four levels: A, B, C, and D.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":3470126069,"compressed_storage_size":3150654899,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/tn-mammo-breast-density/1.0.0/"},{"public_project_uuid":"77aba959-d7eb-4138-9b20-72620312fafe","slug":"mimic-iv-ext-echoqa","version":"1.0.0","core_doi":"10.13026/bxc5-rc50","version_doi":"10.13026/rndk-4s36","is_latest_version":true,"short_description":"We present MIMICEchoQA, a benchmark dataset for echocardiogram-based question answering, built from the publicly available MIMIC-IV-ECHO database.","publish_date":"2025-10-07","title":"MIMIC-IV-ECHO-Ext-MIMICEchoQA: A Benchmark Dataset for Echocardiogram-Based Visual Question Answering","abstract":"We present MIMICEchoQA, a benchmark dataset for echocardiogram-based question answering, built from the publicly available MIMIC-IV-ECHO database. Each echocardiographic study was paired with the closest discharge summary within a 7-day window, and the transthoracic echocardiography (TTE) and ECHO sections were extracted to serve as proxies for cardiologist-authored diagnostic reports. DICOM videos were converted to .mp4 format, and a language model was used to generate multi-turn, closed-ended Q/A pairs grounded in these reports. To ensure anatomical consistency, a view classification model was used to label each video by its echocardiographic view (e.g., A4C, A3C), enabling the filtering of questions referencing structures not visible in the corresponding video. All generated Q/A pairs were manually reviewed by two board-certified cardiologists to ensure clinical validity. The final dataset consists of 622 high-quality, view-consistent Q/A pairs aligned with real-world echocardiogram videos, offering a valuable resource for developing and evaluating models in echocardiographic visual question answering.
Accurate medical decision-making is critical for both patients and clinicians. Patients often find it difficult to interpret their symptoms, determine their severity, and select the right specialist to see. At the same time, clinicians face challenges in integrating complex patient data to make timely, accurate diagnoses. Recent advancements in large language models (LLMs) offer the potential to bridge this gap by supporting decision-making for both patients and healthcare providers. To support the development of LLMs in healthcare, we have curated and extended the MIMIC-IV dataset.
\nWe extracted, processed, and extended the MIMIC-IV dataset, which resulted in 9,150 real-world medical cases with various pathologies. From this larger set 2,200 were further refined to meet the specific needs of our task. This extended dataset is designed to evaluate and improve LLMs' ability to assist with triage, specialist referral, and diagnosis, using critical patient information such as history of present illness, vitals signs and other relevant data. This allows for the development of models that can help both clinicians and patients in making better-informed decisions, ultimately improving healthcare delivery and patient outcomes.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":50637847,"compressed_storage_size":36338392,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimic-iv-ext-cds/1.0.2/"},{"public_project_uuid":"58f11978-c77a-4fbb-96bc-6b4e53108676","slug":"radvlm-model","version":"1.0.0","core_doi":"10.13026/wg4j-wp53","version_doi":"10.13026/50kn-p490","is_latest_version":true,"short_description":"RadVLM is a 7B-parameter vision-language model fine-tuned on public chest-X-ray data that drafts reports, lists abnormalities, grounds findings, and chats about a CXR through a single image-to-text interface.","publish_date":"2025-10-08","title":"RadVLM model","abstract":"We present RadVLM, a compact (7B) multitask conversational foundation model designed for CXR interpretation. Its development relies on the curation of a large-scale instruction dataset comprising over 1 million image-instruction pairs containing both single-turn tasks - such as report generation, abnormality classification, and visual grounding - and multi-turn, multi-task conversational interactions. Our experiments show that RadVLM, fine-tuned on this instruction dataset, achieves state-of-the-art performance in conversational capabilities and visual grounding while remaining competitive in other radiology tasks (report generation, classification). Ablation studies further highlight the benefit of joint training across multiple tasks, particularly for scenarios with limited annotated data. Together, these findings highlight the potential of the RadVLM model as a clinically relevant AI assistant, providing structured CXR interpretation and conversational capabilities to support more effective and accessible diagnostic workflows.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":16077691077,"compressed_storage_size":12382956105,"resource_type":"Model","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/radvlm-model/1.0.0/"},{"public_project_uuid":"40ef241f-379e-4cfa-a129-659ff68211ab","slug":"chexstruct-cxreasonbench","version":"1.0.0","core_doi":"10.13026/wgz6-zx42","version_doi":"10.13026/z3dn-nh22","is_latest_version":false,"short_description":"CheXStruct is an automated pipeline that derives structured diagnostic reasoning steps from chest X-rays. CXReasonBench builds on this to evaluate whether models perform clinically grounded, multi-step reasoning beyond final diagnoses.","publish_date":"2025-10-16","title":"CXReasonBench: A Benchmark for Evaluating Structured Diagnostic Reasoning in Chest X-rays","abstract":"Recent progress in Large Vision-Language Models (LVLMs) has enabled promising applications in medical tasks such as report generation and visual question answering. However, existing benchmarks focus mainly on the final diagnostic answer, offering limited insight into whether models engage in clinically meaningful reasoning.
\nTo address this, we present CheXStruct and CXReasonBench, a structured pipeline and benchmark built on the publicly available MIMIC-CXR-JPG dataset. CheXStruct automatically derives a sequence of intermediate reasoning steps directly from chest X-rays, such as segmenting anatomical regions, deriving anatomical landmarks and diagnostic measurements, computing diagnostic indices, and applying clinical thresholds. CXReasonBench leverages this pipeline to evaluate whether models can perform clinically valid reasoning steps and to what extent they can learn from structured guidance, enabling fine-grained and transparent assessment of diagnostic reasoning. The benchmark comprises 18,988 QA pairs across 12 diagnostic tasks and 1,200 cases, each paired with up to four visual inputs, and supports multi-path, multi-stage evaluation including visual grounding via anatomical region selection and diagnostic measurements.
\nThis dataset is intended to serve as a standardized resource for developing, evaluating, and comparing vision-language models on clinically grounded reasoning tasks in chest X-rays.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":3250182423,"compressed_storage_size":3097678226,"resource_type":"Database","access_policy":"Credentialed","topics":["structured diagnostic pipeline","structured chest x-ray qa","chest x-ray","evaluation","diagnostic reasoning","benchmark","intermediate reasoning steps","grounded reasoning","structured reasoning"],"source_url":"https://physionet.org/content/chexstruct-cxreasonbench/1.0.0/"},{"public_project_uuid":"5e6d62ba-4b48-4a4c-88fa-3cc77802ac8f","slug":"persona-patientsim","version":"1.0.0","core_doi":"10.13026/4b4t-9e66","version_doi":"10.13026/vq0d-v871","is_latest_version":true,"short_description":"PatientSim is a patient simulator that simulates realistic and diverse personas for clinical scenarios, enabling robust training and evaluation of doctor-patient interactions in multi-turn dialogues.","publish_date":"2025-10-18","title":"PatientSim: A Persona-Driven Simulator for Realistic Doctor-Patient Interactions","abstract":"Doctor-patient consultations require multi-turn, context-aware communication tailored to diverse patient personas. Training or evaluating doctor LLMs in such settings requires realistic patient interaction systems. However, existing simulators often fail to reflect the full range of personas seen in clinical practice. To address this, we introduce PATIENTSIM, a patient simulator that generates realistic and diverse patient personas for clinical scenarios, grounded in medical expertise. PATIENTSIM operates using: 1) clinical profiles, including symptoms and medical history, derived from real-world data in the MIMIC-ED and MIMIC-IV datasets, and 2) personas defined by four axes: personality, language proficiency, medical history recall level, and cognitive confusion level, resulting in 37 unique combinations. We evaluated eight LLMs for factual accuracy and persona consistency. The top-performing open-source model, Llama 3.3 70B, was validated by four clinicians to confirm the robustness of our framework. As an open-source, customizable platform, PATIENTSIM provides a reproducible and scalable solution that can be customized for specific training needs. Offering a privacy-compliant environment, it serves as a robust testbed for evaluating medical dialogue systems across diverse patient presentations and shows promise as an educational tool for healthcare.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":155283343,"compressed_storage_size":9887809,"resource_type":"Database","access_policy":"Credentialed","topics":["multi-turn dialogue","llm simulation","electronic health records","doctor-patient consultation"],"source_url":"https://physionet.org/content/persona-patientsim/1.0.0/"},{"public_project_uuid":"c8fa482d-9436-4ba6-8b9e-8940a1b9464a","slug":"chexstruct-cxreasonbench","version":"1.0.1","core_doi":"10.13026/wgz6-zx42","version_doi":"10.13026/d14j-1b45","is_latest_version":true,"short_description":"CheXStruct is an automated pipeline that derives structured diagnostic reasoning steps from chest X-rays. CXReasonBench builds on this to evaluate whether models perform clinically grounded, multi-step reasoning beyond final diagnoses.","publish_date":"2025-10-23","title":"CXReasonBench: A Benchmark for Evaluating Structured Diagnostic Reasoning in Chest X-rays","abstract":"Recent progress in Large Vision-Language Models (LVLMs) has enabled promising applications in medical tasks such as report generation and visual question answering. However, existing benchmarks focus mainly on the final diagnostic answer, offering limited insight into whether models engage in clinically meaningful reasoning.
\nTo address this, we present CheXStruct and CXReasonBench, a structured pipeline and benchmark built on the publicly available MIMIC-CXR-JPG dataset. CheXStruct automatically derives a sequence of intermediate reasoning steps directly from chest X-rays, such as segmenting anatomical regions, deriving anatomical landmarks and diagnostic measurements, computing diagnostic indices, and applying clinical thresholds. CXReasonBench leverages this pipeline to evaluate whether models can perform clinically valid reasoning steps and to what extent they can learn from structured guidance, enabling fine-grained and transparent assessment of diagnostic reasoning. The benchmark comprises 18,988 QA pairs across 12 diagnostic tasks and 1,200 cases, each paired with up to four visual inputs, and supports multi-path, multi-stage evaluation including visual grounding via anatomical region selection and diagnostic measurements.
\nThis dataset is intended to serve as a standardized resource for developing, evaluating, and comparing vision-language models on clinically grounded reasoning tasks in chest X-rays.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":0,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":["structured chest x-ray qa","intermediate reasoning steps","evaluation","structured reasoning","chest x-ray","benchmark","grounded reasoning","diagnostic reasoning","structured diagnostic pipeline"],"source_url":"https://physionet.org/content/chexstruct-cxreasonbench/1.0.1/"},{"public_project_uuid":"589b8b5d-ceec-4045-9731-bb5f62724547","slug":"er-reason","version":"1.0.0","core_doi":"10.13026/jrvj-k081","version_doi":"10.13026/55s7-3c27","is_latest_version":true,"short_description":"The ER-REASON dataset is a longitudinal collection of 25,174 de-identified clinical notes for 3,437 patients admitted to the emergency room (ER) at a large academic medical center between March 1, 2022, and March 31, 2024.","publish_date":"2025-10-23","title":"ER-REASON: A Benchmark Dataset for LLM-Based Clinical Reasoning in the Emergency Room","abstract":"The ER-Reason dataset is a benchmark designed to evaluate LLM-based clinical reasoning and decision-making in the emergency room (ER), a high-stakes setting where clinicians make rapid, consequential decisions across diverse patient presentations and medical specialties under time pressure. This longitudinal collection of de-identified clinical notes encompasses 3,437 patients admitted to the ER at a large academic medical center between March 1, 2022, and March 31, 2024. ER-Reason contains 25,174 notes spanning discharge summaries, progress notes, history and physical exams, consults, echocardiography reports, imaging notes, and ER provider documentation across 3,984 encounters. The benchmark includes evaluation tasks from key stages of the ER workflow: triage intake, initial assessment, treatment selection, disposition planning, and final diagnosis, each structured to reflect core clinical reasoning processes such as differential diagnosis via rule-out reasoning. We also collected 72 full physician-authored rationales explaining reasoning processes that mimic the teaching process used in residency training, and are typically absent from ER documentation. This retrospective dataset captures unstructured, multi-encounter clinical notes reflecting the real-world complexity of ER patient care.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":230862637,"compressed_storage_size":113527339,"resource_type":"Database","access_policy":"Contributor Review","topics":[],"source_url":"https://physionet.org/content/er-reason/1.0.0/"},{"public_project_uuid":"e16da26b-9a4e-4751-9f6a-dfa44262fa80","slug":"heartcycle","version":"1.0.0","core_doi":"10.13026/1cma-6q61","version_doi":"10.13026/z865-eb23","is_latest_version":true,"short_description":"Impedance cardiography dataset (ICG) which combines the ICG signals and other methodologies with the golden standard echocardiographys (ECG). Researchers can use this dataset to compare the ICG points with the real hemodynamic events.","publish_date":"2025-11-03","title":"HeartCycle: A comprehensive dataset of synchronized impedance cardiography and echocardiography for accurate hemodynamic predictions","abstract":"The \"HeartCycle\" dataset offers a comprehensive collection of synchronized impedance cardiography (ICG) and echocardiography (ECHO) signals, supplemented with finger photoplethysmography (PPG), heart sounds, and electrocardiography (ECG) data from 17 healthy volunteers. Collected during the HeartCycle project (FP7-216695), this dataset aims to address biases in the ICG waveform, particularly the ABEXYOZ complex, where the B and X points do not precisely align with the aortic valve opening and closing notches. The biases in B and X point detection are critical for hemodynamic prediction because these characteristic points are used to calculate essential diagnostic parameters, including systolic time intervals (PEP and LVET), contractility, stroke volume, and cardiac output. By providing synchronized ICG and ECHO signals, researchers can better understand these biases and develop more accurate models for hemodynamic parameter computation. The dataset is stored in HDF5 format, facilitating the storage of complex data structures and easy access to various physiological parameters. It is ideal for developing machine learning models to enhance the detection of characteristic points in ICG signals. For instance, machine learning models can be used to detect characteristic points for improved left ventricular ejection time (LVET) estimation or mapping the ICG signal to the different mechanical events in the cardiac cycle using the ECHO as a reference. Detailed metadata and usage notes are included to support data utilization across different software environments. Ethical approval was obtained from the University of Coimbra Hospital's ethics committee, and informed consent was provided by all participants.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":2457906912,"compressed_storage_size":1359816181,"resource_type":"Database","access_policy":"Open","topics":["cardiovascular physiology","electrophysiological study","echocardiography","machine learning","impedance cardiography"],"source_url":"https://physionet.org/content/heartcycle/1.0.0/"},{"public_project_uuid":"01ce2d95-3653-458c-9ee3-3832e36772c2","slug":"medval-bench","version":"1.0.0","core_doi":"10.13026/geme-rz43","version_doi":"10.13026/8ga5-6661","is_latest_version":false,"short_description":"MedVAL-Bench is the first large-scale physician-validated benchmark for medical text validation, spanning 6 diverse medical tasks and containing 840 language model-generated outputs annotated by 12 physicians with error assessments and risk grades.","publish_date":"2025-11-04","title":"MedVAL-Bench: Expert-Annotated Medical Text Validation Benchmark","abstract":"MedVAL-Bench is a dataset containing physician evaluations of errors in language model (LM)-generated medical text. The dataset spans 6 diverse medical text generation tasks and includes annotations from 12 physicians on clinically significant errors for 840 LM-generated outputs. These text-to-text generation tasks involve transforming an input medical text into an output relevant to a specific use case. Each task includes inputs and corresponding LM-generated outputs, which are evaluated for factual consistency by physicians. Importantly, the MedVAL framework and dataset are designed to rely only on inputs for the evaluation process to allow working with datasets that may not have reference outputs, ensuring broad applicability. The evaluation process aims to determine whether the output is factually consistent with the input and is safe for use. MedVAL-Bench constitutes the first large-scale physician-validated benchmark with triage-style risk grading aligned to real-world clinical decision-making, supporting the development of automated, expert-aligned evaluation methods and facilitating research toward trustworthy medical text generation.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1951318,"compressed_storage_size":584121,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/medval-bench/1.0.0/"},{"public_project_uuid":"7ad717bf-2f7a-4300-b30a-deae855dbe90","slug":"hospital-onset-infection","version":"1.0.0","core_doi":"10.13026/pa4z-5293","version_doi":"10.13026/k70x-0m81","is_latest_version":true,"short_description":"NPA-CP is a freely accessible dataset derived from electronic health record (EHR) information at MGB between 2015 and 2024. The dataset includes 11 different pathogens and can be used to predict hospital-onset infections for these pathogens.","publish_date":"2025-11-04","title":"Predictors of Hospital Onset Infection: A Matched Retrospective Cohort Dataset","abstract":"This repository contains a de-identified and curated patient-level dataset for modeling the impact of fine-grained environmental and patient-level factors on nosocomial acquisition of a wide range of drug-susceptible and drug-resistant pathogens. The data are derived from the clinical data warehouses of the Mass General Brigham health system and span all adult inpatients admitted between May 2015 and July 2024.
\nWe use a matched case/control design and provide two datasets reflecting distinctive but complementary strategies to model our outcome. The first, which we call the environmental analysis, investigates the impact of the hospital unit microbiome, using colonization pressure as a proxy, and matches on prior antibiotic exposure to equalize its selective impact across cases and controls. The second strategy, which we call the patient analysis, investigates the impact of prior fine-grained antibiotic exposures and matches controls spatiotemporally to cases based on their time of entry into the same ward, to equalize the impact of the hospital microbiome.
\nEach observation in the dataset corresponds to a hospitalization. We look at nosocomial acquisition of 11 common drug-susceptible and drug-resistant pathogens, with cases having a culture positive for the target pathogen between day 3 and day 30 after admission into their room.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":38336601,"compressed_storage_size":11082414,"resource_type":"Database","access_policy":"Credentialed","topics":["infection control","clinical machine learning","infectious diseases","electronic health records","hospital onset infection","colonization pressure"],"source_url":"https://physionet.org/content/hospital-onset-infection/1.0.0/"},{"public_project_uuid":"e3b0e865-33b0-402e-a21b-7130bdfa3637","slug":"medval-bench","version":"1.0.1","core_doi":"10.13026/geme-rz43","version_doi":"10.13026/653w-3038","is_latest_version":true,"short_description":"MedVAL-Bench is the first large-scale physician-validated benchmark for medical text validation, spanning 6 diverse medical tasks and containing 840 language model-generated outputs annotated by 12 physicians with error assessments and risk grades.","publish_date":"2025-11-14","title":"MedVAL-Bench: Expert-Annotated Medical Text Validation Benchmark","abstract":"MedVAL-Bench is a dataset containing physician evaluations of errors in language model (LM)-generated medical text. The dataset spans 6 diverse medical text generation tasks and includes annotations from 12 physicians on clinically significant errors for 840 LM-generated outputs. These text-to-text generation tasks involve transforming an input medical text into an output relevant to a specific use case. Each task includes inputs and corresponding LM-generated outputs, which are evaluated for factual consistency by physicians. Importantly, the MedVAL framework and dataset are designed to rely only on inputs for the evaluation process to allow working with datasets that may not have reference outputs, ensuring broad applicability. The evaluation process aims to determine whether the output is factually consistent with the input and is safe for use. MedVAL-Bench constitutes the first large-scale physician-validated benchmark with triage-style risk grading aligned to real-world clinical decision-making, supporting the development of automated, expert-aligned evaluation methods and facilitating research toward trustworthy medical text generation.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1951318,"compressed_storage_size":584121,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/medval-bench/1.0.1/"},{"public_project_uuid":"ec87605c-49f2-47c9-bdf1-bbe960b50922","slug":"inredd-dataset-pan924","version":"1.0.0","core_doi":"10.13026/85hv-ct26","version_doi":"10.13026/r5nt-we67","is_latest_version":true,"short_description":"InReDD‑Dataset-V1 is a collection of 924 anonymised panoramic dental radiographs curated by the Interdisciplinary Research Group in Digital Dentistry (InReDD) at the University of São Paulo.","publish_date":"2025-11-23","title":"InReDD-Dataset-PAN924","abstract":"InReDD-Dataset-PAN924 is a collection of 924 radiographic images annotated with mouth and teeth labels by specialists from the InReDD research group.
\nInReDD (Interdisciplinary Research Group in Digital Dentistry) is a collaborative research initiative at the University of São Paulo’s Ribeirão Preto Campus (USP-RP), uniting the Department of Computation and Mathematics (DCM-USP-RP) and the School of Dentistry of Ribeirão Preto (FORP-USP-RP). The group is dedicated to developing applied technologies for the field of Odontology.
\nIn this context, the InReDD-Dataset-PAN924 is an image collection from the field of Odontology. It was developed to support descriptive analyses and to facilitate the creation and validation of artificial intelligence models. The data were collected primarily through clinical work at FORP-RP.
\nThis manuscript draws upon a previously published work, “Development of a dental digital dataset for research in artificial intelligence: the importance of labeling performed by radiologists.” However, certain details have been adjusted or updated to account for temporal adaptations and contextual revisions. As a result, portions of the content may not correspond verbatim to the original publication, although the scientific essence and core contributions remain preserved.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":1019485994,"compressed_storage_size":982458853,"resource_type":"Database","access_policy":"Contributor Review","topics":[],"source_url":"https://physionet.org/content/inredd-dataset-pan924/1.0.0/"},{"public_project_uuid":"7c070c4b-ac5f-4bd0-a0ae-c941d8c41420","slug":"echograph-note2num-annotations","version":"1.0.0","core_doi":"10.13026/2h1h-9n57","version_doi":"10.13026/hb5q-9532","is_latest_version":true,"short_description":"EchoGraph is a model that automatically extracts and structures clinical information from echocardiogram reports. The Annotated ECHO-NOTE2NUM Dataset contains MIMIC-III echo reports enhanced with EchoGraph annotations to enhance future research.","publish_date":"2025-12-04","title":"EchoGraph-annotated ECHO-NOTE2NUM examples","abstract":"This repository releases the EchoGraph-annotated ECHO-NOTE2NUM dataset, containing 45,794 echocardiography reports with comprehensive entity and relation annotations. Each report from the ECHO-NOTE2NUM dataset has been automatically annotated using EchoGraph, a BERT-based information extraction model specifically designed for echocardiography reports. The annotations employ a tailored schema capturing clinical observations (definitely present, definitely absent, uncertain), anatomical structures, measurements, and four types of relations (Modify, Gauge, Located at, Suggestive of). The dataset contains a total of 1,709,074 entities and 671,512 relations across all 45,794 reports. EchoGraph was developed using 600 densely annotated Mayo Clinic reports (48,256 entities, 29,731 relations) and validated on 60 MIMIC-EchoNote reports, demonstrating strong performance (entity F1 0.85 internal, 0.80 external; relation F1 0.70 internal, 0.52 external).
\nThis annotated dataset enables research in clinical NLP, automated report evaluation, and development of AI systems for echocardiography.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":458626759,"compressed_storage_size":33581483,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/echograph-note2num-annotations/1.0.0/"},{"public_project_uuid":"488814d7-76df-4b0d-84c9-d4e629b4188a","slug":"armd-mgb","version":"1.0.0","core_doi":"10.13026/53vq-f537","version_doi":"10.13026/2r5k-b955","is_latest_version":true,"short_description":"ARMD-MGB contains detailed microbiology and clinical metadata for >225,000 patients and >970,000 cultures collected over 10 years","publish_date":"2025-12-05","title":"Antibiotic Resistance Microbiology Dataset Mass General Brigham (ARMD-MGB)","abstract":"The Antibiotic Resistance Microbiology Dataset – MGB (ARMD-MGB) is a de-identified resource derived from electronic health records (EHR) that facilitates research in antimicrobial resistance (AMR). ARMD-MGB encompasses data collected from over 225,000 adult patients over 10 years from hospitals in the Mass General Brigham healthcare system. The focus of the data is on over 970,000 microbiological cultures, with associated antibiotic susceptibilities, and clinical and demographic features of the patients who submitted the samples. Key attributes include organism identity, semi-quantitative antibiotic susceptibility results, susceptibility phenotypes, and de-identified clinical metadata.
\nThe ARMD-MGB dataset is designed to complement the ARMD-Stanford dataset. Cohort inclusion and exclusion criteria, feature descriptions, and encoding are similar across sites. These datasets support studies on antimicrobial stewardship, causal inference, and clinical decision-making, and are designed to be reusable and interoperable, promoting collaboration and innovation in combating AMR.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":6573000423,"compressed_storage_size":415100009,"resource_type":"Database","access_policy":"Credentialed","topics":["antimicrobial resistance","electronic health records","medical informatics"],"source_url":"https://physionet.org/content/armd-mgb/1.0.0/"},{"public_project_uuid":"29ff2588-62ae-4a33-8d99-1cf6d3d3c8d0","slug":"micro-nav","version":"1.1.1","core_doi":"10.13026/tyj5-dd44","version_doi":"10.13026/rtc6-cq72","is_latest_version":true,"short_description":"This study explores the respiratory microbiome's role in nosocomial lower respiratory tract infections in ICU patients. Conducted in Chía, Colombia, revealing the microbiome's impact on disease progression.","publish_date":"2025-12-05","title":"Microbiological, Immunological and Biochemical Characteristics of the Development of Ventilator Associated Pneumonia","abstract":"The respiratory microbiome plays a critical role in metabolism, immune system maturation, and protection against pathogens. Traditionally, respiratory microbiology in pneumonia focused on identifying a specific pathogen, often disregarding normal oral flora as contaminants. However, recent research highlights the importance of the pulmonary microbiome's altered composition (dysbiosis) and its association with diseases such as Ventilator-Associated Pneumonia (VAP). The relationship between dysbiosis and immune dysregulation remains unclear, raising questions about whether dysbiosis is a cause or consequence of disease progression. This dataset explores the respiratory microbiome's role in nosocomial lower respiratory tract infections in ICU patients.
\nThis prospective, observational dataset was built in an ICU in Chia, Colombia, focusing on adult patients requiring mechanical ventilation without initial pneumonia. Clinical data and biological samples, including bronchoalveolar lavage, orotracheal secretion, and blood, were collected at multiple time points. Microbiome analysis involved DNA extraction, gene sequencing of the 16S ribosomal unit, and bioinformatic analysis to characterize microbial diversity. Concurrently, pulmonary, and systemic inflammatory markers were quantified using ELISA.
\nThe dataset enrolled 141 patients, generating a comprehensive dataset on demographic, clinical, and microbiome variables. This research not only provides insights into the microbial ecology of VAP but also establishes a repository of data for future studies. The findings underscore the complex interplay between the microbiome and disease in critically ill patients, with implications for improving clinical outcomes in the ICU setting.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":5185883,"compressed_storage_size":405191,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/micro-nav/1.1.1/"},{"public_project_uuid":"792f76c7-cc90-440d-a34b-a523bb667be1","slug":"medvh","version":"1.0.1","core_doi":"10.13026/wjx8-sx53","version_doi":"10.13026/0qtp-3d10","is_latest_version":true,"short_description":"MedVH provides a visual hallucination evaluation benchmark for large language models in the medical context. It formulates tests using chest X-ray images, including multi-choice question answering and long-text generation tasks.","publish_date":"2025-12-10","title":"MedVH: Towards Systematic Evaluation of Hallucination for Large Vision Language Models in the Medical Context","abstract":"Large Vision Language Models (LVLMs) have recently achieved superior performance in various tasks on natural image and text data, which inspires a large amount of studies for LVLMs fine-tuning and training. Despite their advancements, there has been scant research on the robustness of these models against hallucination when fine-tuned on smaller datasets. In this study, we introduce a new benchmark dataset, the Medical Visual Hallucination evaluation benchmark (MedVH), to evaluate the hallucination of domain-specific LVLMs. MedVH comprises six tasks to evaluate hallucinations in LVLMs within the medical context, which includes tasks for a comprehensive understanding of textual and visual input, as well as long textual response generation.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1475671088,"compressed_storage_size":1470005250,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/medvh/1.0.1/"},{"public_project_uuid":"7eaee72f-5efb-4236-b581-0bf5cd412dbe","slug":"b2ai-voice","version":"3.0.0","core_doi":"10.13026/37yb-1t42","version_doi":"10.13026/k81f-qr68","is_latest_version":false,"short_description":"A dataset of features from voice recordings and metadata to enable the development, benchmarking, and validation of clinically applicable machine-learning models for diagnosing a wide range of health conditions.","publish_date":"2025-12-16","title":"Bridge2AI-Voice: An ethically-sourced, diverse voice dataset linked to health information","abstract":"The human voice contains complex acoustic markers which have been linked to important health conditions including dementia, mood disorders, and cancer. When viewed as a biomarker, voice is a promising characteristic to measure as it is simple to collect, cost-effective, and has broad clinical utility. Recent advances in artificial intelligence have provided techniques to extract previously unknown prognostically useful information from dense data elements such as images. The Bridge2AI-Voice project seeks to create an ethically sourced flagship dataset to enable future research in artificial intelligence and support critical insights into the use of voice as a biomarker of health. Here we present Bridge2AI-Voice, a comprehensive collection of data derived from voice recordings with corresponding clinical information.
\nBridge2AI-Voice v3.0 contains data for 833 participants across five sites in North America. Participants were selected based on known conditions which manifest within the voice waveform including voice disorders, neurological disorders, mood disorders, and respiratory disorders. The release contains data considered low risk, including derivations such as spectrograms but not the original voice recordings. Detailed demographic, clinical, and validated questionnaire data are also made available.
","license":{"name":"Bridge2AI Voice Registered Access License"},"dua":{"name":"Bridge2AI Voice Registered Access Agreement"},"main_storage_size":13879111829,"compressed_storage_size":13810977689,"resource_type":"Database","access_policy":"Credentialed","topics":["bridge2ai","voice"],"source_url":"https://physionet.org/content/b2ai-voice/3.0.0/"},{"public_project_uuid":"cec9c34f-37c8-4d89-ac4d-4950dd81a4c7","slug":"b2ai-voice-pediatric","version":"1.0.0","core_doi":"10.13026/mf9s-5r03","version_doi":"10.13026/y7mp-eh56","is_latest_version":false,"short_description":"A dataset of questionnaire responses, spectrograms, and other information for pediatric participants collected for the Bridge2AI voice as a biomarker of health project.","publish_date":"2025-12-17","title":"Bridge2AI-Voice Pediatric Dataset","abstract":"The human voice contains complex acoustic markers which have been linked to important health conditions including dementia, mood disorders, and cancer. When viewed as a biomarker, voice is a promising characteristic to measure as it is simple to collect, cost-effective, and has broad clinical utility. Recent advances in artificial intelligence have provided techniques to extract previously unknown prognostically useful information from dense data elements such as images. The Bridge2AI-Voice project seeks to create an ethically sourced flagship dataset to enable future research in artificial intelligence and support critical insights into the use of voice as a biomarker of health. Here we present Bridge2AI-Voice, a comprehensive collection of data derived from voice recordings with corresponding clinical information.
\nBridge2AI-Voice Pediatric Dataset v3.0 contains derived audio features for 22,620 recordings collected from 300 participants aged 2-18. The release contains data considered low risk, including derivations such as spectrograms but not the original voice recordings. Detailed demographic, clinical, and validated questionnaire data are also made available.
","license":{"name":"Bridge2AI Voice Registered Access License"},"dua":{"name":"Bridge2AI Voice Registered Access Agreement"},"main_storage_size":3118959438,"compressed_storage_size":2904560736,"resource_type":"Database","access_policy":"Credentialed","topics":["voice","bridge2ai"],"source_url":"https://physionet.org/content/b2ai-voice-pediatric/1.0.0/"},{"public_project_uuid":"22586be7-3b1f-41ac-92fd-cf80325e877d","slug":"archehr-qa-bionlp-task-2025","version":"1.3","core_doi":"10.13026/zzax-sy62","version_doi":"10.13026/n708-sn25","is_latest_version":true,"short_description":"A dataset for grounded question answering (QA) from electronic health records (EHRs).","publish_date":"2026-01-02","title":"ArchEHR-QA: A Dataset for Addressing Patient's Information Needs related to Clinical Course of Hospitalization","abstract":"Patient’s unique information needs about their hospitalization can be addressed using clinical evidence from electronic health records (EHRs) and artificial intelligence (AI). However, robust datasets to assess the factuality and relevance of AI-generated responses are lacking and, to our knowledge, none capture patient information needs in the context of their EHRs. To address this gap, we introduce ArchEHR-QA, an expert-annotated dataset of 134 cases from intensive care unit and emergency department settings to evaluate the grounding capabilities of models for responding to patient-initiated queries. The dataset consists of patient-initiated questions posted in public domain, the corresponding clinician-interpreted questions, the excerpts of the EHRs annotated at the sentence-level with relevance to the question, and clinician-generated free-text answers to the questions grounded with EHR sentences. We collect true patient health information needs expressed in real-world health forum messages, we then align the messages to publicly accessible real EHRs. To our knowledge, this is the first public dataset that encapsulates patient questions and relevant clinical evidence from EHRs. We further provide an evaluation framework to assess two critical aspects of a grounded EHR QA system: does it identify relevant information from given clinical evidence and does it use this information in responding to user queries.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1155892,"compressed_storage_size":214535,"resource_type":"Challenge","access_policy":"Credentialed","topics":["question answering","electronic health record","patient portals","clinicians"],"source_url":"https://physionet.org/content/archehr-qa-bionlp-task-2025/1.3/"},{"public_project_uuid":"7dfd8f30-9279-4973-9e44-60ac83f269a9","slug":"psg-ipa","version":"1.0.0","core_doi":"10.13026/7kvm-qe34","version_doi":"10.13026/esx0-nw71","is_latest_version":true,"short_description":"The HMC-IPA dataset comprises 20 PSG recordings, each with manual and computer-assisted scorings by 12 sleep technologists, for studying inter-scorer variability and evaluating automated sleep analysis algorithms","publish_date":"2026-01-08","title":"PSG-IPA: A PolySomnoGraphic Inter-scorer Performance Assessment database","abstract":"This dataset, named PolySomnoGraphic Inter-scorer Performance Assessment (PSG-IPA) database, comprises 20 polysomnography (PSG) recordings and related multi-expert scoring annotations. It was utilized in a study to evaluate inter-scorer variability, specifically assessing reliability and scoring times associated with various PSG review tasks. The study also aimed to compare the impact of computer-assisted (semi-automatic) scoring versus unassisted visual analysis on these factors.
\nThe dataset includes recordings of common raw PSG signal activity, such as electroencephalography (EEG), electrooculography (EOG), electromyography (EMG) of both the chin and lower limbs, electrocardiography (ECG), and various respiratory derivations. Each recording is accompanied by both manual and computer-assisted scorings, performed by 12 independent expert sleep technologists. This comprehensive resource offers significant value for further research into PSG inter-scorer variability, as well as for the development and performance assessment of automated sleep stage and event detection algorithms.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":2749550640,"compressed_storage_size":2250875014,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/psg-ipa/1.0.0/"},{"public_project_uuid":"2257c23e-154e-4dbd-9f4f-0876e12513c3","slug":"lunguage","version":"1.0.0","core_doi":"10.13026/t075-g517","version_doi":"10.13026/pk42-4v91","is_latest_version":true,"short_description":"A radiologist-annotated benchmark of structured chest X-ray reports at single and sequential levels, comprising 1,473 reports across 18 relation types and 80 longitudinal cases.","publish_date":"2026-01-12","title":"Lunguage: A Benchmark for Structured and Sequential Chest X-ray Interpretation","abstract":"Radiology reports convey detailed clinical observations and capture diagnostic reasoning that evolves over time. However, existing evaluation methods are limited to single-report settings and rely on coarse metrics that fail to capture fine-grained clinical semantics and temporal dependencies. We introduce LUNGUAGE, a benchmark dataset of structured radiology reports that serves as a gold standard for evaluating structured report frameworks. It is designed to support comprehensive assessment of both single-report interpretation and longitudinal reasoning. Constructed from a subset of the MIMIC-CXR test set, LUNGUAGE comprises 1,473 chest X-ray reports from 230 patients, annotated with over 17,000 expert-verified entities and 23,000 relation–attribute pairs across 18 relation types. An additional subset of 80 sequential reports from 10 patients captures disease progression across 3 to 14 studies per patient, covering time intervals from 1 to 1,200 days. These are annotated with over 41,000 pairwise comparisons, grouped into semantically and temporally coherent groups. The dataset also includes a schema-aligned vocabulary covering diagnostic entities and attributes. All annotations were conducted and verified by board-certified radiologists, resulting in a clinically grounded resource for structured understanding and temporal reasoning in radiology.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":17656579,"compressed_storage_size":1034332,"resource_type":"Database","access_policy":"Credentialed","topics":["fine-grained structured reports","attribute-level clinical reasoning","medical text structuring","longitudinal clinical reasoning","chest x-ray report parsing","medical information structuring","benchmark dataset for radiology report","medical information extraction","structured radiology reports","temporal relation extraction","radiology report benchmarking","longitudinal clinical understanding"],"source_url":"https://physionet.org/content/lunguage/1.0.0/"},{"public_project_uuid":"9948a418-3fea-4f06-b4c8-ac01687d0e47","slug":"snomed-ct-entity-challenge","version":"1.2.0","core_doi":"10.13026/3f69-tk39","version_doi":"10.13026/swfn-tz92","is_latest_version":false,"short_description":"272 discharge notes from the MIMIC-IV-Note dataset annotated with SNOMED CT concepts.","publish_date":"2026-01-12","title":"SNOMED CT Entity Linking Challenge","abstract":"This challenge, sponsored by SNOMED International, seeks to advance the development of Entity Linking models that operate on unstructured clinical texts. Participants in the challenge will train entity linking models using a subset of MIMIC-IV-Note discharge summaries that have been annotated with SNOMED CT concepts by a team of medical professionals. The full dataset (which is comprised of a training set and a test set) consists of approximately 75,000 annotations across 272 discharge summaries.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":7366654,"compressed_storage_size":2220333,"resource_type":"Challenge","access_policy":"Credentialed","topics":["clinical annotation","snomed","entity linking"],"source_url":"https://physionet.org/content/snomed-ct-entity-challenge/1.2.0/"},{"public_project_uuid":"afcba537-9de4-47d5-9834-aa8bb99b64c4","slug":"vet-diagnosis-coding","version":"1.0.0","core_doi":"10.13026/e5w8-1018","version_doi":"10.13026/akca-bb83","is_latest_version":true,"short_description":"Fine-tuned GatorTron LLM for veterinary diagnosis coding to 7,739 SNOMED-CT codes based on clinical summary text from the Colorado State University Veterinary Teaching Hospital.","publish_date":"2026-01-25","title":"Fine-tuning foundational models to code diagnoses from veterinary health records","abstract":"Veterinary medical records represent a large data resource for application to veterinary and One Health clinical research efforts. Use of the data is limited by interoperability challenges including inconsistent data formats and data siloing. Clinical coding using standardized medical terminologies enhances the quality of medical records and facilitates their interoperability with veterinary and human health records from other sites. Previous studies, such as DeepTag and VetTag, evaluated the application of Natural Language Processing (NLP) to automate veterinary diagnosis coding, employing long short-term memory (LSTM) and transformer models to infer a subset of Systematized Nomenclature of Medicine - Clinical Terms (SNOMED-CT) diagnosis codes from free-text clinical notes. This study expands on these efforts by incorporating all 7,739 distinct SNOMED-CT diagnosis codes recognized by the Colorado State University (CSU) Veterinary Teaching Hospital (VTH) and by leveraging the increasing availability of pre-trained language models (LMs). Twelve freely-available pre-trained LMs (GatorTron, MedicalAI ClinicalBERT, medAlpaca, VetBERT, PetBERT, BERT, BERT Large, RoBERTa, GPT-2, GPT-2 XL, DeBERTa V3, and ModernBERT) were fine-tuned on the free-text notes from 246,473 manually-coded veterinary patient visits included in the CSU VTH's electronic health records (EHRs), which resulted in superior performance relative to previous efforts. The most accurate results were obtained when expansive labeled data were used to fine-tune relatively large clinical LMs, but the study also showed that comparable results can be obtained using more limited resources and non-clinical LMs. The results of this study contribute to the improvement of the quality of veterinary EHRs by investigating accessible methods for automated coding and support both animal and human health research by paving the way for more integrated and comprehensive health databases that span species and institutions.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":15757113806,"compressed_storage_size":13086948640,"resource_type":"Model","access_policy":"Credentialed","topics":["foundational models","one health","diagnoses","snomed-ct","veterinary medicine","large language models","omop cdm","veterinary medical records","transformers","natural language processing","clinical coding"],"source_url":"https://physionet.org/content/vet-diagnosis-coding/1.0.0/"},{"public_project_uuid":"02bb8da8-bd9b-4d7a-a61b-3020c8f7e600","slug":"olst-mocap-forceplate-radar","version":"1.0","core_doi":"10.13026/70xa-vm53","version_doi":"10.13026/46hn-6b25","is_latest_version":true,"short_description":"A multimodal dataset of 32 participants performing the One-Legged Stand Test (OLST), with synchronized motion capture, force plate, and 24 GHz radar data. Each of 1,241 trials is labeled with foot-lift, stability phases, and foot-touchdown.","publish_date":"2026-01-26","title":"Multimodal Synchronized Motion Capture, Force Plate, and Radar Dataset of the One-Legged Stand Test for Fall-Risk Assessment","abstract":"We present synchronized motion capture, force plate, and processed 24 GHz Frequency-Modulated Continuous Wave (FMCW) radar recordings from 32 healthy participants (15 “young” participants aged 32 or younger, and 17 “old” participants aged 64 or older) performing the One-Legged Stand Test (OLST), a clinically validated fall risk assessment. The defining events of each of the 1,241 OLST attempts (foot-lift, start-of-stability, end-of-stability, and foot-touchdown) were labeled across data types using synchronized motion capture and force plate recordings. The dataset supports research in biomechanics, digital biomarker development, radar signal processing, and fall risk characterization. Basic MATLAB and Python utilities are provided for data loading, visualization, and event alignment.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":16161531924,"compressed_storage_size":10922641870,"resource_type":"Database","access_policy":"Open","topics":["human movement","fall risk assessment","non-contact sensing","one-legged stand test","motion capture","force plate analysis","digital biomarkers","human balance testing","geriatrics","radar signal processing","postural control","human pose estimation","multimodal sensing","biomechanics","aging and mobility"],"source_url":"https://physionet.org/content/olst-mocap-forceplate-radar/1.0/"},{"public_project_uuid":"ca4c2f88-902d-452e-a6d0-9f1137f61688","slug":"tollet","version":"1.0.1","core_doi":"10.13026/z36n-vk32","version_doi":"10.13026/411k-1476","is_latest_version":true,"short_description":"We present tOLIet, the first thigh ECG dataset with real signals captured by a toilet seat with electrodes. There are 149 recordings from 86 people, useful for research into cardiovascular assessment using \"invisible\" ECG.","publish_date":"2026-02-02","title":"tOLIet: Single-lead Thigh-based Electrocardiography Using Polimeric Dry Electrodes","abstract":"Our team previously introduced an innovative concept for an \"invisible\" Electrocardiography (ECG) system, incorporating electrodes and sensors into a toilet seat design to enable signal acquisition from the thighs. Building upon that work, we now present a novel dataset featuring real-world, single-lead ECG signals captured at the thighs, offering a valuable resource for advancing research on thigh-based ECG for cardiovascular disease assessment. To our knowledge, this is the first dataset of its kind.
\nThe tOLIet dataset comprises 149 ECG recordings collected from 86 individuals (50 females, 36 males) with an average age of 31.73 ± 13.11 years, a mean weight of 66.89 ± 10.70 kg, and an average height of 166.82 ± 6.07 cm. Participants were recruited through direct contact with the Principal Investigator at Centro Hospitalar Universitário de Lisboa Central (CHULC) and via clinical consultations conducted at the same institution. Each recording includes four differential signals acquired from electrode pairs embedded in the toilet seat, with reference signals obtained from a standard 12-lead hospital ECG system.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":545360267,"compressed_storage_size":78232764,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/tollet/1.0.1/"},{"public_project_uuid":"4055827a-c2e7-41c7-bc42-9039fa88db67","slug":"brugada-huca","version":"1.0.0","core_doi":"10.13026/5pfd-cs22","version_doi":"10.13026/0m2w-dy83","is_latest_version":true,"short_description":"Brugada syndrome is a rare but potentially life-threatening cardiac arrhythmia disorder, with an elevated risk of sudden cardiac death. This dataset introduces 12-lead ECG recordings gather to support the study of this rare disease.","publish_date":"2026-02-03","title":"Brugada-HUCA: 12-Lead ECG Recordings for the Study of Brugada Syndrome","abstract":"We present a novel dataset of 12-lead electrocardiogram (ECG) recordings developed to support the study and classification of Brugada syndrome, a rare but potentially fatal cardiac arrhythmia. The data were collected retrospectively from patients evaluated at the Cardiology Department of the Hospital Universitario Central de Asturias (HUCA) and reviewed by clinical experts. Diagnostic labels were assigned according to internationally accepted criteria.
\nThe dataset comprises 363 subjects in total, including 76 patients diagnosed with Brugada syndrome and 287 healthy control subjects. Each recording is accompanied by diagnostic information.
\nThis resource aims to facilitate the development and validation of machine learning algorithms and computational tools for ECG interpretation, supporting improved early detection and clinical decision-making. The dataset, fully anonymized in accordance with ethical and privacy standards, provides a reproducible foundation for collaborative research into Brugada syndrome and related cardiac conditions.
","license":{"name":"Creative Commons Attribution-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":18717603,"compressed_storage_size":15121344,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/brugada-huca/1.0.0/"},{"public_project_uuid":"daf595fb-98db-451a-a62f-bc3462150856","slug":"mimic-iii-ext-ppg","version":"1.0.0","core_doi":"10.13026/fcvm-qs78","version_doi":"10.13026/nmwb-6h34","is_latest_version":false,"short_description":"Large-Scale, Quality-Assessed PPG-based Benchmark Dataset for Cardiovascular and Respiratory Signal Analysis based on MIMIC-III","publish_date":"2026-02-05","title":"MIMIC-III-Ext-PPG: A PPG Benchmark Dataset for Cardiorespiratory Analysis","abstract":"MIMIC-III-Ext-PPG is a large-scale, annotated, and quality-assessed dataset of photoplethysmogram (PPG) segments extracted from the MIMIC-III Waveform Database Matched Subset. It contains 4.9 million non-overlapping 30-second segments from 6,131 critically-ill patients and serves as the largest public resource for heart rhythm classification. The dataset contains signals recorded during several heart rhythms: sinus rhythm, atrial fibrillation, atrial flutter, heart block, pacing, and several less common rhythms. PPG segments were extracted within the 15-minute period prior to a patient's heart rhythm being recorded in the medical record, providing reference rhythm annotations in-line with best practice from the literature. In addition, simultaneous electrocardiogram (ECG, lead II), arterial blood pressure (ABP), and respiratory (RESP) signals are provided where available. Additional annotations were derived from these signals, including blood pressure, heart rate, and respiratory rate. Signal quality metrics are provided for all signals. Furthermore, a dedicated pipeline was applied to remove mislabeled signals, which, to the best of our knowledge, have previously gone unnoticed. The MIMIC-III-Ext-PPG dataset is a valuable resource for a variety of downstream prediction tasks, enabled by its size, range of reference annotations, and signal quality metrics.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":124983504782,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":["heart rhythm","photoplethysmography","signal quality","critical care","electrocardiogram","blood pressure","respiratory rate"],"source_url":"https://physionet.org/content/mimic-iii-ext-ppg/1.0.0/"},{"public_project_uuid":"3c3addda-e366-4233-b28d-c472a2a21eda","slug":"snomed-ct-entity-challenge","version":"1.2.1","core_doi":"10.13026/3f69-tk39","version_doi":"10.13026/az93-vm52","is_latest_version":true,"short_description":"272 discharge notes from the MIMIC-IV-Note dataset annotated with SNOMED CT concepts.","publish_date":"2026-02-17","title":"SNOMED CT Entity Linking Challenge","abstract":"This challenge, sponsored by SNOMED International, seeks to advance the development of Entity Linking models that operate on unstructured clinical texts. Participants in the challenge will train entity linking models using a subset of MIMIC-IV-Note discharge summaries that have been annotated with SNOMED CT concepts by a team of medical professionals. The publicly available portion of this dataset consists of approximately 75,000 annotations across 272 discharge summaries.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":7366654,"compressed_storage_size":2220333,"resource_type":"Challenge","access_policy":"Credentialed","topics":["entity linking","snomed","clinical annotation"],"source_url":"https://physionet.org/content/snomed-ct-entity-challenge/1.2.1/"},{"public_project_uuid":"88579cb4-8a37-4a6e-89eb-ac51e85c401a","slug":"vitaldb-arrhythmia","version":"1.0.0","core_doi":"10.13026/cprk-3z36","version_doi":"10.13026/axd6-wm13","is_latest_version":true,"short_description":"We present a comprehensive intraoperative arrhythmia dataset with 734,528 seconds of ECG recordings from 482 patients, featuring over 660,000 beats annotated and validated by five anesthesiologists.","publish_date":"2026-02-27","title":"VitalDB Arrhythmia Database: An Anesthesiologist-Validated Large-Scale Intraoperative Arrhythmia Dataset with Beat and Rhythm Labels","abstract":"Intraoperative cardiac arrhythmias present distinct characteristics and clinical challenges compared to non-surgical environments, yet publicly available electrocardiogram (ECG) databases have primarily focused on ambulatory and intensive care environments. To address this gap, we present the VitalDB Arrhythmia Database, a comprehensive collection of annotated intraoperative ECG recordings specifically designed for developing and validating arrhythmia detection algorithms in the surgical context. The database comprises 734,528 seconds of continuous ECG data from 482 surgical patients, with over 660,000 individually annotated heartbeats classified across four beat types and 10 distinct rhythm categories. To efficiently process the extensive source data, we developed a custom deep learning beat classifier that served as an automated screening tool for arrhythmia candidate segments. All annotations underwent rigorous validation by five anesthesiologists, with each segment independently reviewed by at least two anesthesiologists. Inter-rater reliability analysis demonstrated excellent agreement with an overall Cohen's kappa of 0.930 ± 0.130. This publicly accessible resource provides the research community with clinically validated intraoperative arrhythmia data, facilitating the development of robust detection algorithms suited to the unique physiological and technical challenges of the perioperative environment.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":21926799,"compressed_storage_size":3534362,"resource_type":"Database","access_policy":"Open","topics":["vitaldb","ppg","arterial waveform","ecg","intraoperative dataset"],"source_url":"https://physionet.org/content/vitaldb-arrhythmia/1.0.0/"},{"public_project_uuid":"e5988ab0-e3d1-4a71-8a7d-bd0e0987f918","slug":"mimic-iv-echo-ext-lvvol-a4c","version":"1.0.0","core_doi":"10.13026/ft1n-4y69","version_doi":"10.13026/713s-z339","is_latest_version":true,"short_description":"A curated subset of MIMIC-IV-ECHO providing apical four-chamber cine loops with manual ROI masks, volumetric labels, and ready-to-use MP4/NPZ derivatives for robust LV volume and ejection fraction research.","publish_date":"2026-02-27","title":"MIMIC-IV-ECHO-Ext-LVVOLUMES-A4C-ROI: Annotated Subset of Apical Four-Chamber Echocardiography for PoCUS-Style LV Volume and Function Analysis","abstract":"MIMIC-IV-ECHO-Ext-LVVOLUMES-A4C-ROIis a curated, credentialed subset of MIMIC-IV-ECHO v0.1 focusing on apical four-chamber (A4C) echocardiographic views with accompanying left-ventricular volumetric labels. The release includes 1,064 de-identified A4C DICOM video sequences from 806 unique patients, each paired with a manually annotated region-of-interest (ROI) mask in both PNG and JSON formats at the original DICOM 4D resolution. ROI-applied, downsampled (256 × 256) MP4 and NPZ derivatives are provided for efficient model development. The full-resolution masks preserve native geometry, enabling regeneration of higher-resolution datasets directly from the source DICOMs.
\nThis resource supports three primary use cases: (i) as an out-of-distribution (OOD) PoCUS external test set for evaluating state-of-the-art LV volume and ejection fraction (EF) models trained on laboratory echocardiography; (ii) training and benchmarking ROI segmentation models for background and overlay removal; and (iii) developing LV-focused video models using preprocessed MP4/NPZ inputs that eliminate DICOM parsing overhead.
\nBoth single-plane (A4C) and biplane (BP) volumetric and EF labels are included, enabling transparent comparison between monoplane interpretability and biplane geometric precision. Access is provided under the PhysioNet Credentialed Health Data License (v1.5.0).
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":10673285801,"compressed_storage_size":9870212039,"resource_type":"Database","access_policy":"Credentialed","topics":["lvesv","roi segmentation","cardiac video analysis","left ventricular volume","mimic-iv-echo","apical four-chamber","quantitative cardiology","biplane simpson","transformer models","lvef","ejection fraction","echocardiography","deep learning","a4c","pocus","ultrasound","lvedv","medical imaging","domain adaptation","dicom"],"source_url":"https://physionet.org/content/mimic-iv-echo-ext-lvvol-a4c/1.0.0/"},{"public_project_uuid":"9efa634b-6a83-49ea-a82f-09d0b5b23f45","slug":"neuro-stress-resilience-hci","version":"1.0.0","core_doi":"10.13026/9m4y-3n35","version_doi":"10.13026/x3vc-p627","is_latest_version":true,"short_description":"This dataset contains multimodal neurophysiological and physiological recordings collected from participants performing cognitively demanding tasks to study the temporal dynamics of stress resilience during human-computer interaction","publish_date":"2026-02-27","title":"Neurophysiological Dataset of Stress Resilience During Human-Computer Interaction","abstract":"This dataset provides multimodal neurophysiological and physiological recordings collected to investigate stress resilience. It comprises electroencephalography, functional near-infrared spectroscopy, electrodermal activity, blood volume pulse, inter-beat intervals, heart rate, and accelerometer readings from 35 participants engaged in a computer-based task under stress. The experimental design consisted of six sequential phases: Resting Baseline, Working Baseline, Stress 1, Recovery 1, Stress 2, and Recovery 2. Event markers were synchronized across all modalities to track transitions between task conditions. Additionally, subjective resilience was assessed using the 10-item Connor-Davidson Resilience Scale. The dataset also includes demographic variables such as age, gender, and ethnic background. This dataset is valuable for research on stress resilience, neuroergonomics, and machine learning applications in human factors. Researchers can utilize the data to analyze neural and physiological responses to stress, develop predictive models, and evaluate adaptive training strategies. By making this dataset publicly available, we aim to support reproducibility, interdisciplinary collaboration, and advancements in neuroergonomics.
","license":{"name":"Open Data Commons Open Database License v1.0"},"dua":null,"main_storage_size":20459129828,"compressed_storage_size":4979989467,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/neuro-stress-resilience-hci/1.0.0/"},{"public_project_uuid":"29a4fcec-d057-47b8-b579-03dbd6c06dbd","slug":"mimic-iii-ext-notes","version":"1.0.0","core_doi":"10.13026/2kv6-1s83","version_doi":"10.13026/9tfx-yx07","is_latest_version":true,"short_description":"We evaluated general large language models' performance in clinical information extraction on MIMIC-III notes.","publish_date":"2026-02-27","title":"MIMIC-III-Ext-Notes","abstract":"Unstructured clinical documentation, such as progress notes, contains rich contextual information critical for clinical decision-making but remains underutilized in computational research due to the limited availability of annotated datasets. The MIMIC-III-Ext-Notes dataset was developed to address this gap by providing a resource for evaluating large language models (LLMs) and other natural language processing (NLP) systems in extracting and contextualizing clinical information.
\nThe dataset includes 150 clinical notes randomly sampled from the MIMIC-III Clinical Database, from which 2,288 clinical concepts were identified using MetaMap and annotated by clinicians for detection accuracy, encounter relevance, and negation status. The resulting dataset enables the evaluation of models’ abilities to recognize, interpret, and reason about symptom mentions and disease concepts within realistic clinical narratives. By incorporating both concept-level and contextual annotations, MIMIC-III-Ext-Notes provides a valuable benchmark for developing, testing, and validating NLP and LLM frameworks designed for clinical text understanding and decision support applications.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":601542,"compressed_storage_size":202763,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimic-iii-ext-notes/1.0.0/"},{"public_project_uuid":"59880d3c-2068-4efd-95bc-9347cac85f51","slug":"respiratory-oximetry-apnoea","version":"1.0.0","core_doi":"10.13026/60xk-ky06","version_doi":"10.13026/s45r-k263","is_latest_version":true,"short_description":"This dataset contains airway pressure, flow and pulse oximetry waveforms from 20 healthy adults during simulated apnoea events, including arterial and venous PPG signals for developing and validating OSA detection and oxygenation models.","publish_date":"2026-03-05","title":"Respiratory and Pulse Oximetry Waveforms from Healthy Adults During Simulated Apnoea Events","abstract":"This dataset contains airway gauge pressure, inspiratory differential pressure, and pulse oximetry waveforms collected from 20 healthy adult participants (10 females, 10 males) during simulated obstructive sleep apnoea (OSA) trials. Recordings include photoplethysmography (PPG) signals from a custom neck-worn reflectance oximeter, which records both arterial and venous blood waveforms from the carotid artery and internal jugular vein, respectively. Each subject completed a a series of 8 × 60-second breathing trials with varying positive airway pressures (0, 4, and 8 cmH₂O) set with a continuous positive airway pressure (CPAP) device. Normal breathing was recorded initially before subsequent trials of three breaths followed by 10- and 20-second breath holds, before normal breathing for the rest of the 60 seconds. The dataset is structured to support algorithm development and validation for OSA detection, modelling of airway resistance and oxygen desaturation responses, and non-invasive estimation of arterial and venous oxygen saturation. Supplementary code and figures are included to facilitate data analysis.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":250749434,"compressed_storage_size":39291925,"resource_type":"Database","access_policy":"Open","topics":["respiratory","pulse oximetry","obstructive sleep apnea"],"source_url":"https://physionet.org/content/respiratory-oximetry-apnoea/1.0.0/"},{"public_project_uuid":"fe3f81e4-2e22-4827-9cd3-50da2a08096e","slug":"mouse-stroke-cylinder-rearing","version":"1.0.0","core_doi":"10.13026/tyn4-2a75","version_doi":"10.13026/8bnm-6e37","is_latest_version":true,"short_description":"This dataset provides cylinder rearing video and behavioral scoring data from 59 mice undergoing stroke and drug treatments.","publish_date":"2026-03-05","title":"Longitudinal Cylinder Rearing Behavioral Data in a Mouse Stroke Model Across Multiple Drug Treatments","abstract":"This dataset includes longitudinal data from 59 mice assigned to four drug treatment groups (lemborexant 10 mg/kg, lemborexant 30 mg/kg, zolpidem 30 mg/kg, and vehicle), assessed at five timepoints: baseline (pre-treatment), maximal deficit (post-stroke 48 hours), and weeks 2, 4, and 8 post-treatment. Designed to evaluate drug efficacy over time, the dataset captures functional or physiological changes across stroke acute and recovery phases via the behavioral test of cylinder rearing. The shared data specifically comprises 295 video recordings (.mp4) of these sessions, paired with corresponding frame-based annotation files (.csv) that classify paw usage (left, right, both, or neither) for each frame.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":22656258358,"compressed_storage_size":22638084728,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/mouse-stroke-cylinder-rearing/1.0.0/"},{"public_project_uuid":"8d69d449-04cb-441b-8a6f-dd41468ff447","slug":"mimic-iv-echo","version":"1.0","core_doi":"10.13026/agyq-zp57","version_doi":"10.13026/nrjh-5r77","is_latest_version":false,"short_description":"The MIMIC-IV-ECHO module contains structured measurements from over 200,000 echocardiograms and more than 500,000 echocardiogram DICOM files. Patients overlap with those in the MIMIC-IV Clinical Database.","publish_date":"2026-03-10","title":"MIMIC-IV-ECHO: Echocardiogram Matched Subset","abstract":"The MIMIC-IV-ECHO module contains structured echocardiographic measurements and DICOM files from echocardiography exams.
\nStructured measurements are available from 206,488 echocardiogram studies (including 179,928 transthoracic, 16,389 stress, and 10,171 transesophageal echocardiograms). These are from 91,372 unique patients who are in the MIMIC-IV Clinical Database and who had echocardiograms conducted between 2008 – 2022. The dataset encompasses a broad range of cardiologist-recorded echocardiographic measurements, including chamber sizes and volumes, systolic and diastolic function, valvular morphology and function, aortic root and great vessel dimensions, and Doppler-derived hemodynamics. Unstructured measurements are present in the cardiologist’s free-text reports, which will be made available later under the MIMIC-IV-Note module.
\nThe echocardiogram DICOMs contain more than 500,000 DICOMs across 7,243 studies from 4,579 distinct patients. This subset contains transthoracic echocardiograms for patients who appear in the MIMIC-IV Clinical Database and were admitted between 2017–2019. A given study consists of numerous sequences of images, with each sequence representing a particular view of the patient's heart.
\nRecords in MIMIC-IV-ECHO are matched to the MIMIC-IV Clinical Database, making it possible to link to information across the MIMIC-IV modules. We have also provided information for linking the DICOMs to the structured measurements and notes where available.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1896740561762,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimic-iv-echo/1.0/"},{"public_project_uuid":"d18348b4-985a-4cea-8969-613fdcd2fc2c","slug":"mimic-iii-ext-ca","version":"1.0.0","core_doi":"10.13026/5a4t-7324","version_doi":"10.13026/ec0n-y377","is_latest_version":true,"short_description":"The MIMIC-III-Ext-CA dataset contains annotations of 31 PPG-captured cardiac arrest episodes from the MIMIC-III clinical and waveform databases.","publish_date":"2026-03-11","title":"MIMIC-III-Ext-CA: a MIMIC-III Derived Dataset of Cardiac Arrests in Photoplethysmographs","abstract":"Photoplethysmography (PPG) is increasingly considered for detecting out-of-hospital cardiac arrest (OHCA), but publicly available datasets remain scarce. We used a method for identifying cardiac arrest episodes captured by PPG in the MIMIC-III database, which combines automated screening with manual annotation of waveforms and clinical data. Using this approach, we compiled 36 annotated cardiac arrest episodes from 31 patients. The dataset serves as a valuable resource for developing and validating wearable OHCA detection technologies.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":328359,"compressed_storage_size":295127,"resource_type":"Database","access_policy":"Credentialed","topics":["cardiac arrest","out-of-hospital cardiac arrest","mimic-iii","photoplethysmography","ppg","ohca"],"source_url":"https://physionet.org/content/mimic-iii-ext-ca/1.0.0/"},{"public_project_uuid":"40deac33-f432-4984-b8e0-21409822d6c6","slug":"hillel-yaffe-glaucoma-dataset","version":"1.1.0","core_doi":"10.13026/pdxv-m215","version_doi":"10.13026/m92s-0z95","is_latest_version":true,"short_description":"HYGD is a rigorously annotated fundus image dataset with gold-standard clinical labels designed to improve and benchmark deep learning models for accurate glaucoma detection.","publish_date":"2026-03-17","title":"Hillel Yaffe Glaucoma Dataset (HYGD): A Gold-Standard Annotated Fundus Dataset for Glaucoma Detection","abstract":"Glaucomatous optic neuropathy (GON) is a leading cause of irreversible blindness worldwide, affecting an estimated 64.3 million people globally with projections reaching 111.8 million by 2040. Approximately 50% of cases remain undiagnosed until advanced stages when vision loss becomes noticeable. Traditional diagnosis requires comprehensive ophthalmic examinations by specialists, creating accessibility barriers in many regions.
\nThe Hillel Yaffe Glaucoma Dataset (HYGD) addresses a critical limitation in existing GON datasets: the lack of gold-standard annotations. Unlike most publicly available datasets where glaucoma labels are determined solely from digital fundus images (DFIs), HYGD's labels are based on comprehensive ophthalmic examinations, including visual acuity assessment, intraocular pressure measurement, optical coherence tomography (OCT), visual field tests, and at least one year of follow-up monitoring.
\nThe dataset is structured to include both the DFIs and a labels file containing patient IDs, GON classifications, and image quality scores. All DFIs were taken using a TOPCON DRI OCT Triton retinal camera with a 45° FOV and underwent deidentification and standardization processing.
\nHYGD enables researchers to train and benchmark models on rigorously annotated data, potentially improving their reliability. This dataset serves as a valuable resource for developing generalizable models that can function across diverse patient populations and clinical settings, ultimately supporting earlier detection and treatment of GON.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":131865715,"compressed_storage_size":124214721,"resource_type":"Database","access_policy":"Open","topics":["ophthalmology","glaucoma","dfi","gon","fundus","gold-standard","retina"],"source_url":"https://physionet.org/content/hillel-yaffe-glaucoma-dataset/1.1.0/"},{"public_project_uuid":"6bd02e5d-7c31-4ce9-bce7-b79acfc8134d","slug":"mimic-iii-ext-ppg","version":"1.1.0","core_doi":"10.13026/fcvm-qs78","version_doi":"10.13026/r6k1-xt76","is_latest_version":true,"short_description":"Large-Scale, Quality-Assessed PPG-based Benchmark Dataset for Cardiovascular and Respiratory Signal Analysis based on MIMIC-III","publish_date":"2026-03-17","title":"MIMIC-III-Ext-PPG: A PPG Benchmark Dataset for Cardiorespiratory Analysis","abstract":"MIMIC-III-Ext-PPG is a large-scale, annotated, and quality-assessed dataset of photoplethysmogram (PPG) segments extracted from the MIMIC-III Waveform Database Matched Subset. It contains around 6.3 million non-overlapping 30-second segments from 6,189 critically-ill patients and serves as the largest public resource for heart rhythm classification. The dataset contains signals recorded during several heart rhythms: sinus rhythm, atrial fibrillation, atrial flutter, heart block, pacing, and several less common rhythms. PPG segments were extracted within the 15-minute period prior to a patient's heart rhythm being recorded in the medical record, providing reference rhythm annotations in line with best practice from the literature. In addition, simultaneous electrocardiogram (ECG, lead II), arterial blood pressure (ABP), and respiratory (RESP) signals are provided where available. Additional annotations were derived from these signals, including blood pressure, heart rate, and respiratory rate. Signal quality metrics are provided for all signals. The MIMIC-III-Ext-PPG dataset is a valuable resource for a variety of downstream prediction tasks, enabled by its size, range of reference annotations, and signal quality metrics.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":160525455472,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":["critical care","blood pressure","electrocardiogram","signal quality","heart rhythm","photoplethysmography","respiratory rate"],"source_url":"https://physionet.org/content/mimic-iii-ext-ppg/1.1.0/"},{"public_project_uuid":"9f5b692d-5c8e-4346-9bf2-f4280c4dd1ae","slug":"echoview","version":"0.1","core_doi":"10.13026/av0c-x376","version_doi":"10.13026/ywz0-5b62","is_latest_version":true,"short_description":"ECHOVIEW provides structured viewing class annotations for 29,196 transthoracic echocardiograms derived from MIMIC-IV-ECHO using a pretrained CNN. Manual clinician review shows substantial agreement (κ=0.69) with these annotations.","publish_date":"2026-03-17","title":"Structured Viewing Classification Annotations From the MIMIC-IV-ECHO Dataset (ECHOVIEW)","abstract":"The development of robust echocardiography analyses has been limited by the scarcity of well-annotated, labeled echocardiogram datasets. We introduce the Structured Viewing Classification Annotations From the MIMIC-IV-ECHO Dataset (ECHOVIEW), a derived dataset providing transthoracic (TTE) echocardiogram viewing classifications for a derived subset of the MIMIC-IV-ECHO dataset. This derived dataset includes viewing classifications for 29,196 echocardiograms across 717 studies, from 463 patients. Each study contains multiple echocardiogram views and sometimes uses multiple ultrasound techniques, with changes in the position and angle of the ultrasound probe corresponding to different views. Likewise, each patient can have multiple studies. The ECHOVIEW dataset features 23 separate viewing classifications predicted for each echocardiogram, which are produced by a machine learning (ML) classifier. Our model provides classification probabilities for echocardiograms with at least 10 frames of data. We validated our predictions by having a licensed cardiologist and an intensivist independently review a sample of 50 echocardiograms. Their evaluations were then compared against the ML view classifications. With the release of this dataset, we provide machine-generated view classifications for the MIMIC-IV-ECHO dataset and hope to encourage the development of further echocardiography analyses and models.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":4129137,"compressed_storage_size":696215,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/echoview/0.1/"},{"public_project_uuid":"48e8e74c-3a31-481d-a0e1-7241d74acba3","slug":"mimic-iv-ext-medicalbench","version":"1.0.0","core_doi":"10.13026/0k22-xc18","version_doi":"10.13026/j98m-g356","is_latest_version":true,"short_description":"This dataset is an evidence‑grounded benchmark built on MIMIC‑IV discharge summaries that evaluates how well large language models can verify ICD‑10 medical concepts, including implicitly documented diagnoses, by identifying supporting text evidence.","publish_date":"2026-03-24","title":"MIMIC-IV-Ext-MedicalBench: Evaluating Large Language Models Towards Improved Medical Concept Extraction","abstract":"Medical concept extraction from electronic health records underpins many downstream applications, yet remains challenging because medically meaningful concepts, such as diagnosis, are frequently implied rather than explicitly stated in medical narratives. Existing benchmarks with human-annotated evidence spans underscore the importance of grounding extracted concepts in medical text. However, they predominantly focus on explicitly stated concepts and provide limited coverage of cases in which medically relevant concepts must be inferred. We present MedicalBench, a new benchmark for medical concept extraction with evidence grounding that evaluates implicit medical reasoning. MedicalBench formulates concept extraction as a verification task over medical note–concept pairs, coupled with sentence-level evidence identification. Built from MIMIC-IV discharge summaries and human-verified ICD-10 codes, the dataset is curated through a multi-stage large language model (LLM) triage pipeline followed by dual medical annotation and expert review. It deliberately includes implicit positives, semantically confusable negatives, and cases where LLM judgments disagree with human assessments. Annotators provide sentence-level evidence spans and concise medical rationales. In total, the dataset contains 405 high-quality examples, covering a broad range of ICD-10 chapters. By providing ground-truth evidence and confusable alternatives, MedicalBench enables rigorous evaluation of not only whether a model can extract the correct concept, but also why — rewarding solutions that can highlight relevant evidence and reject plausible-but-incorrect diagnosis and procedures.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":202896,"compressed_storage_size":70029,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimic-iv-ext-medicalbench/1.0.0/"},{"public_project_uuid":"e7eb7ea3-bbb6-477f-ad61-d6026adbc115","slug":"mimic-iv-ext-pe","version":"1.0.0","core_doi":"10.13026/f77d-vr35","version_doi":"10.13026/5qq1-zv65","is_latest_version":true,"short_description":"CTPA (computed tomography pulmonary angiogram) radiology reports from MIMIC-IV with pulmonary embolism (PE) adjudication","publish_date":"2026-03-24","title":"MIMIC-IV-Ext-PE: Pulmonary Embolism Labels for CT Pulmonary Angiography Radiology Reports","abstract":"Pulmonary embolism (PE) is a leading cause of preventable in-hospital mortality. Advances in diagnosis, risk stratification, and prevention can improve outcomes. Large, publicly available datasets are needed to move research forward but are lacking in the field of hemostasis and thrombosis. In this study, we added PE labels to computed tomography pulmonary angiography (CTPA) radiology reports in MIMIC-IV. We used Regular Expression (RegEx) to identify CTPA radiology reports (n=19,942) and extracted sentences containing PE-related words (“snippets”). Two physicians manually reviewed these snippets, referring to the full report as needed, and labeled each report as PE positive or negative. Positive labels included any acute PE (n=1,591). Acute PE that only involved subsegmental arteries were labeled as subsegmental. Negative labels included chronic PE, equivocal findings, and no PE (n=18,351). Using this as a gold standard, we then compared the performance of a finetuned transformer model to diagnosis codes in their ability to classify the reports as PE positive or negative.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":57268890,"compressed_storage_size":12775659,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimic-iv-ext-pe/1.0.0/"},{"public_project_uuid":"b2b7ad46-c5cb-4d84-aa15-bff9f8a17645","slug":"mimic-iv-ext-clif","version":"1.1.0","core_doi":"10.13026/86jk-m086","version_doi":"10.13026/j481-g420","is_latest_version":true,"short_description":"Transforming the MIMIC-IV 3.1 database into the Common Longitudinal ICU data Format (CLIF)","publish_date":"2026-03-24","title":"MIMIC-IV-Ext-CLIF: MIMIC-IV in the Common Longitudinal ICU data Format (CLIF)","abstract":"MIMIC-IV-Ext-CLIF is a derived dataset of MIMIC-IV v3.1, transformed into the Common Longitudinal ICU data Format (CLIF). CLIF is an open-source critical care data model developed by a consortium of more than 10 US academic medical centers to standardize intensive care unit (ICU) data for multi-center research. While CLIF has demonstrated value in federated research settings, access has been limited to consortium members with institutional electronic health record (EHR) data. This dataset addresses that gap by providing CLIF-formatted data derived from the freely accessible de-identified MIMIC-IV dataset, enabling researchers worldwide to utilize the CLIF format without requiring institutional EHR access.
\nMIMIC-IV-Ext-CLIF contains 14 CLIF tables as of the latest CLIF 2.1.0 version, covering core demographics (patient, hospitalization), clinical monitoring (vitals, labs), respiratory support, medication administration, and specialized ICU interventions (e.g., continuous renal replacement therapy). The transformation employs a reproducible ETL pipeline with transparent mapping decisions documented in user-friendly spreadsheets.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1215130648,"compressed_storage_size":982736410,"resource_type":"Database","access_policy":"Credentialed","topics":["mimic","clif","the common longitudinal icu data format","critical care"],"source_url":"https://physionet.org/content/mimic-iv-ext-clif/1.1.0/"},{"public_project_uuid":"0e27cca1-e70b-49fd-ad2c-039c7888556c","slug":"clinical-trajectory-flow-icu","version":"1.0.0","core_doi":"10.13026/ey1p-0p43","version_doi":"10.13026/pqe3-bv16","is_latest_version":true,"short_description":"This resource comprises three clinical time series datasets used in the paper Trajectory Flow Matching with Applications to Clinical Time Series Modeling to evaluate models for handling irregularly sampled data in critical care settings.","publish_date":"2026-03-24","title":"Clinical Time Series Datasets for Trajectory Flow Matching Evaluation: ICU Sepsis, ICU Cardiac Arrest, and ICU GIB Cohorts","abstract":"This resource comprises three clinical time series datasets used in the paper Trajectory Flow Matching with Applications to Clinical Time Series Modeling to evaluate models for handling irregularly sampled data in critical care settings. The ICU Sepsis and ICU Cardiac Arrest datasets are derived from the eICU Collaborative Research Database v2.0, and the ICU Gastrointestinal Bleeding (GIB) Dataset is derived from the MIMIC-III database. The ICU Sepsis Dataset includes 3362 patients with sepsis as the primary admission diagnosis, the ICU Cardiac Arrest Dataset contains 64589 ICU patients at risk for cardiac arrest, and the ICU GIB Dataset includes 2602 patients with gastrointestinal bleeding. Each dataset is split into training, validation, and test sets with a ratio of 0.8/0.1/0.1. These datasets were used to demonstrate the improved performance of the Trajectory Flow Matching (TFM). Sharing these datasets aims to facilitate further exploration of clinical time series modeling, the development of new models in the medical domain, and to help others replicate the experimental results.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":222570970,"compressed_storage_size":27952400,"resource_type":"Database","access_policy":"Credentialed","topics":["clinical time series"],"source_url":"https://physionet.org/content/clinical-trajectory-flow-icu/1.0.0/"},{"public_project_uuid":"4266f0dd-239f-4612-8b5d-6bec52e88270","slug":"digital-data-ingestion-utility","version":"1.0.0","core_doi":"10.13026/gmhb-9z50","version_doi":"10.13026/jt51-cs74","is_latest_version":true,"short_description":"Schema-only database for a patient's digital signal data (e.g., ECG) collection using a novel or investigational device. Key features include real-time linkage between digital physiological recordings and their corresponding metadata.","publish_date":"2026-03-25","title":"Digital Data Real-Time Ingestion Utility","abstract":"We present a Python-based utility for real-time ingestion of digital physiological signals and study enrollment metadata into an Oracle database. The application combines filesystem event monitoring, a GUI data entry form, and thin-mode Oracle connectivity to ensure timely and accurate data capture at the point of file creation. The system is designed for research assistants to enter metadata for digital signal recordings (e.g., ECG, PCG, and other physiological datasets), with automated database field assignments (e.g., study_date, study_id) and compliance with pre-defined enumerations and constraints. The utility follows best practices and aligns with high institutional and regulatory standards for digital data quality and integrity, ensuring robust, reproducible, and traceable research data collection.
\nKey features include real-time linkage between digital physiological recordings and their corresponding metadata, ensuring synchronized and traceable data entry. The use of the Oracle thin client enables lightweight and efficient database integration, while the GUI-based control of data input minimizes human errors during metadata collection. All entries are securely timestamped to maintain a comprehensive audit trail and support full accountability. From a security standpoint, database credentials should be stored in environment variables or .env files with restricted access, user accounts should be limited to INSERT privileges, and read-only database views are recommended for downstream data consumers to preserve data integrity and confidentiality.
","license":{"name":"MIT License"},"dua":null,"main_storage_size":29621,"compressed_storage_size":14366,"resource_type":"Software","access_policy":"Open","topics":["empty database","reliability","bedside digital signal collection","quality control","schema-only database","real-time recording"],"source_url":"https://physionet.org/content/digital-data-ingestion-utility/1.0.0/"},{"public_project_uuid":"2c51cee9-4f93-4b3c-b08d-07345d5c0922","slug":"mimic-cxr-ext-ils","version":"1.0.0","core_doi":"10.13026/gdq0-ej95","version_doi":"10.13026/8ejy-4t06","is_latest_version":true,"short_description":"Instruction-guided lesion segmentation data for chest X-rays, including 1.1M instruction-answer pairs and 91K segmentation masks covering seven major lesion types.","publish_date":"2026-03-25","title":"MIMIC-CXR-Ext-ILS: Lesion Segmentation Masks and Instruction-Answer Pairs for Chest X-rays","abstract":"The applicability of current lesion segmentation models for chest X-rays (CXRs) has been limited both by a small number of target labels and the reliance on complex, expert-level text inputs, creating a barrier to practical use. To address these limitations, we introduce instruction-guided lesion segmentation (ILS), a medical-domain adaptation of referring image segmentation (RIS) designed to segment diverse lesion types based on simple, user-friendly instructions. Under this task, we construct MIMIC-CXR-Ext-ILS, the first large-scale instruction-answer dataset for CXR lesion segmentation, using our fully automated multimodal pipeline that generates annotations from CXR images and their corresponding reports. MIMIC-CXR-Ext-ILS contains 1.1M instruction-answer pairs derived from 192K images and 91K unique segmentation masks, covering seven major lesion types. Despite being constructed entirely without human intervention, expert evaluations report a high acceptance rate of over 95% for this dataset.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":929981912,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":["text-guided segmentation","lesion segmentation","chest x-ray segmentation"],"source_url":"https://physionet.org/content/mimic-cxr-ext-ils/1.0.0/"},{"public_project_uuid":"2ac82519-a152-4757-947a-f62e5f92c376","slug":"script-carpediem-dataset","version":"1.9.0","core_doi":"10.13026/1qke-qx33","version_doi":"10.13026/etpr-0n55","is_latest_version":false,"short_description":"SCRIPT contains clinical data and physician adjudication from patients admitted to an intensive care unit, received mechanical ventilation, and underwent bronchoalveolar lavage sampling because of pneumonia suspicion.","publish_date":"2026-03-26","title":"SCRIPT CarpeDiem Dataset: demographics, outcomes, and per-day clinical parameters for critically ill patients with suspected pneumonia","abstract":"Traditional approaches to analyzing episodes of patient care in the ICU examine features on presentation and outcomes on discharge, collapsing the numerous events that happen during a patient’s stay and ignoring intercurrent ICU complications or interventions. With CarpeDiem, we aimed to examine clinical features on a per-day basis as done during the common practice of daily physician rounds. The Successful Clinical Response in Pneumonia Therapy (SCRIPT) CarpeDiem Dataset features 21,931 patient-hospital-days from 704 patients enrolled in the SCRIPT study between June 2018 and May 2023, all of whom were critically ill patients on mechanical ventilation suspected of having pneumonia who underwent a bronchoalveolar lavage as part of routine clinical care. Each patient has demographics, pneumonia episode information as adjudicated by a panel of critical care physicians, outcomes, and 113 clinical parameters for which data are present including vital signs, laboratory parameters, and mechanical support devices. Data have been deidentified per Health Insurance Portability and Accountability Act (HIPAA) Safe Harbor rules. This dataset combines expert clinician adjudication with per-day granular information to provide a unique tool to examine the clinical courses of critically ill patients with pneumonia. We use this dataset in a manuscript examining the contribution of ventilator-associated pneumonia to patient outcomes (see Usage section for details) and share it for others to work with.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":38776963,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/script-carpediem-dataset/1.9.0/"},{"public_project_uuid":"a38cf150-2e23-4d26-a4c9-30f027413da4","slug":"embedding-brset-mbrset","version":"1.0.0","core_doi":"10.13026/evz5-0948","version_doi":"10.13026/1h4p-vz70","is_latest_version":true,"short_description":"Precomputed image embeddings for the BRSET and mBRSET Brazilian retinal datasets to support efficient, secure, and equitable ophthalmic AI research, enabling tasks such as classification, clustering, multimodal modeling, and fairness analysis.","publish_date":"2026-03-30","title":"Embedding-Based Representations for BRSET and mBRSET","abstract":"BRSET and mBRSET are publicly available Brazilian ophthalmological datasets composed of curated retinal fundus photographs with associated clinical and demographic information. While these resources enable diverse research applications, training deep learning models directly on high-resolution images is computationally intensive and often restricted by privacy regulations limiting the circulation of identifiable medical images. To address these challenges and facilitate equitable reuse, this project provides a comprehensive release of precomputed image embeddings for both datasets. These representations were generated using state-of-the-art vision backbones: DINOv3 ViT-S/16 (384-d) and ViT-B/16 (768-d) for transformer-based features, alongside ConvNeXt-Tiny (768-d) and ConvNeXt-Base (1024-d) for convolutional features. All models were applied in inference-only mode with a standardized preprocessing pipeline. Each fundus photograph was converted into a fixed-length numerical vector and exported as a CSV file, where each row corresponds to a single image and its respective embedding. These representations preserve critical semantic and structural information, enabling downstream tasks such as clustering, similarity search, multimodal modeling, disease classification, and fairness assessment without requiring raw pixel access. By providing scalable, privacy-preserving embeddings derived from Brazilian ophthalmic data, this resource reduces computational barriers, accelerates AI model development, and supports global research participation, particularly in low-resource environments, ensuring that advanced ophthalmic AI tools are accessible to a broader scientific community.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":460495759,"compressed_storage_size":205218007,"resource_type":"Database","access_policy":"Credentialed","topics":["computer vision","ophthalmology","vector embeddings"],"source_url":"https://physionet.org/content/embedding-brset-mbrset/1.0.0/"},{"public_project_uuid":"1260ee8d-428c-44f3-a3e6-8f82640dedde","slug":"big-ideas-glycemic-wearable","version":"1.1.3","core_doi":"10.13026/w591-tp72","version_doi":"10.13026/aw6y-fc44","is_latest_version":true,"short_description":"Glucose measurements and wrist-worn wearable sensor data from highnormoglycemic participants.","publish_date":"2026-04-13","title":"BIG IDEAs Lab Glycemic Variability and Wearable Device Data","abstract":"This study aimed to determine the feasibility and effectiveness of wearable devices in detecting early physiological changes prior to the development of prediabetes [1-3]. The study generated digital biomarkers for remote, mHealth-based prediabetes and hyperglycemia risk to classify which individuals should undergo further clinical testing. The primary inclusion criteria were subjects aged 35-65 years, inclusive, including only post-menopausal females, with a point of care A1C measurement between 5.2-6.4%, inclusive. Blood was collected during the study for measurement of glucose, hemoglobin A1C, lipoproteins, and triglycerides. Participants wore a Dexcom 6 continuous glucose monitor (CGM) and an Empatica E4 wristband for 10 days while receiving a standardized breakfast meal every other day. At the end of the 10 days, the participant returned to the clinic for an oral glucose tolerance test (OGTT). Research data collected includes physiological measurements from wearable devices such as heart rate, accelerometry, and electrodermal conductance.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":36645551880,"compressed_storage_size":5015250233,"resource_type":"Database","access_policy":"Open","topics":["biomedical engineering","pre-diabetes","biomarkers"],"source_url":"https://physionet.org/content/big-ideas-glycemic-wearable/1.1.3/"},{"public_project_uuid":"0378df2a-3197-434d-b582-798b7198e9e8","slug":"inspire","version":"1.4","core_doi":"10.13026/dv8j-5412","version_doi":"10.13026/7fje-8522","is_latest_version":false,"short_description":"A public dataset that contains information related to surgery, anesthesia, laboratory results, medications, diagnosis, and outcomes from 50% of the patients who received surgery at Seoul National University Hospital between 2011 and 2020.","publish_date":"2026-04-16","title":"INSPIRE, a publicly available research dataset for perioperative medicine","abstract":"We present the INSPIRE dataset, a publicly available research dataset in perioperative medicine, which includes approximately 130,000 cases (50% of all surgical cases) who underwent anesthesia for surgery at an academic institution in South Korea between 2011 and 2020. This comprehensive dataset includes patient characteristics such as age, sex, American Society of Anesthesiologists physical status classification, diagnosis, surgical procedure code, department, and type of anesthesia. It also includes vital signs in the operating theatre, general wards, and intensive care units (ICUs), laboratory results from six months before admission to six months after discharge, and medication during hospitalization. Complications include total hospital and ICU length of stay and in-hospital death. We hope this dataset will inspire collaborative research and development in perioperative medicine and serve as a reproducible external validation dataset to improve surgical outcomes.
","license":{"name":"Korea Credentialed Health Data License-1-0-0"},"dua":{"name":"Korea Credentialed Health Data Agreement-1-0-0"},"main_storage_size":1068848881,"compressed_storage_size":1055130897,"resource_type":"Database","access_policy":"Credentialed","topics":["open dataset","multi-center","surgery","perioperative medicine"],"source_url":"https://physionet.org/content/inspire/1.4/"},{"public_project_uuid":"d44fc56c-24c1-4abd-86b9-e17248150bb6","slug":"grabmyo-flow","version":"1.0.0","core_doi":"10.13026/4e7j-qa41","version_doi":"10.13026/ah52-zq51","is_latest_version":true,"short_description":"GRABMyoFlow extends the GRABMyo sEMG dataset with dynamic transitions and 20 new subjects (63 total). Features multi-day wrist recordings for robust hand gesture recognition and biometrics research.","publish_date":"2026-04-16","title":"GRABMyoFlow - Dataset extension","abstract":"This dataset is an extension of the Gesture Recognition and Biometrics Electromyogram (GRABMyo) dataset, a comprehensive collection of surface electromyography (sEMG) signals from the forearm and wrist while performing static gestures, designed for hand gesture recognition (HGR) and biometric authentication research. The study extends the dataset in two key aspects: 1) by adding wrist-only recordings from 20 newly recruited healthy adults, and 2) by including a dynamic trial at the end of each session. The extension study follows the protocol established in the original study to ensure methodological consistency and reliable data acquisition. Combined with prior wrist recordings, the extended dataset now includes data from 63 participants, significantly increasing the sample size. Signal strength and quality in the new recordings have been validated to match the standards of the original dataset, ensuring uniformity across all subjects. The dynamic gesture recordings included in the extension capture the natural transitions between hand poses, better simulating real-world movement scenarios. These dynamic data enable new avenues for continuous authentication and gesture-event segmentation, directly addressing prior limitations of static-only datasets. Data acquisition employed the established electrode configuration, and open access to the data in the waveform database (WFDB)-format is provided to support the widespread use of research. This enriched dataset enhances both the scale and temporal complexity of the GRABMyo dataset, delivering a robust resource for advancing sEMG-based HGR for human-computer interaction or biometric systems research.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":4725931979,"compressed_storage_size":4591106365,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/grabmyo-flow/1.0.0/"},{"public_project_uuid":"95952578-72fe-4cb8-8f04-5706b6c34dc5","slug":"ecg-pacing-artifact-removal","version":"1.0.0","core_doi":"10.13026/xd1x-zt03","version_doi":"10.13026/hq2h-8p89","is_latest_version":true,"short_description":"The software application (written in MATLAB) automatically detects pacing spikes (pacing artifacts) on a 12-lead ECG signal and automatically removes them. User-friendly GUI permits reviewing the results and adjusting settings to achieve perfection.","publish_date":"2026-04-24","title":"Automated Detection and Removal of Pacing Artifacts from ECG Signal","abstract":"Pacing artifacts in electrocardiogram (ECG) signals can interfere with accurate waveform analysis, making their detection and removal essential for reliable interpretation. We present an algorithm and provide MATLAB code to detect and remove pacing artifacts from raw digital 12-lead ECG recordings. The algorithm was developed using the Chronic Renal Insufficiency Cohort (CRIC) data, which includes 493 paced ECGs, and validated on an external dataset (273 paced ECGs). A graphical user interface (GUI) was designed to allow user-friendly quality control of automated algorithm performance. Users can adjust settings to achieve the desired result: complete removal of the pacing spike without removing the ECG waveform. The algorithm automatically detected and removed pacing artifacts in 82.8% of the 10-second ECG files in the validation set, while the remaining files required user input. The overall sensitivity and specificity for pacing artifact detection were 98.4% and 76.5%, respectively. In summary, the proposed algorithm effectively detects and removes ECG pacing artifacts with high automation, thereby minimizing the need for manual fine-tuning.
","license":{"name":"MIT License"},"dua":null,"main_storage_size":1022275,"compressed_storage_size":982725,"resource_type":"Software","access_policy":"Open","topics":["cardiac pacing","artifact","digital signal processing","electrocardiogram"],"source_url":"https://physionet.org/content/ecg-pacing-artifact-removal/1.0.0/"},{"public_project_uuid":"38746e32-e8b8-4f83-9d80-6315029dd4ad","slug":"ki-endolist","version":"1.0.0","core_doi":"10.13026/84gy-nv92","version_doi":"10.13026/k99q-fm63","is_latest_version":true,"short_description":"This database contains daily symptoms of 34 endometriosis patients over 1-10 months of monitoring. It includes basic patient information, frequency and intensity of symptoms, and standard MedDRA symptom mapping for clinical interpretation.","publish_date":"2026-04-30","title":"KI EndoLIST: Endometriosis Longitudinal Individualized Symptoms Tracking Dataset","abstract":"Endometriosis affects approximately 10% of reproductive-age women globally, yet the time to diagnosis is four to twelve years due to clinical challenges and the normalization of symptoms by patients and healthcare providers. Additionally, highly diverse symptom profiles of patients lead to suboptimal treatment approaches and prolonged patient suffering.
\nThis unique database addresses the critical gap in curating individual endometriosis symptoms longitudinally. Unlike periodical standardized questionnaires, our custom-developed app allowed each of 34 Israeli endometriosis patients to document their unique disease burden daily, using individualized symptom sets and severity scales.
\nThe dataset includes an onboarding patient information file, an onboarding code dictionary, per-user longitudinal daily symptom monitoring data (and a corresponding data dictionary), and standardized mapping of symptoms to the Medical Dictionary for Regulatory Activities (MedDRA) for clinical interpretation. It enables dynamic evaluation of symptom variability, severity, and individual disease complexity at the patient level.
\nThis dataset represents a valuable resource for researchers and clinicians seeking to understand the true complexity of endometriosis symptom experience. It enables examination of personalized symptom patterns and severity, with the purpose of promoting optimized clinician-patient engagement and individualized treatment approaches. By capturing the nuanced reality of living with endometriosis, this dataset can inform more effective diagnostic strategies and management protocols tailored to individual patient needs.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":910878,"compressed_storage_size":160100,"resource_type":"Database","access_policy":"Restricted","topics":[],"source_url":"https://physionet.org/content/ki-endolist/1.0.0/"},{"public_project_uuid":"357b9f72-68b0-471f-9be9-175381f73ece","slug":"multimodal-gait-dataset","version":"1.0.0","core_doi":"10.13026/f7ez-gr08","version_doi":"10.13026/r0ea-7161","is_latest_version":true,"short_description":"Dataset of gait recordings from 59 healthy adults, combining brain activity, muscle activity, body kinematics, and ground forces during treadmill walking at three different speeds.","publish_date":"2026-04-30","title":"A multimodal gait dataset of brain activity, muscle activity, kinematics and ground forces in young adults","abstract":"Gait is a fundamental motor function, and its analysis is essential for understanding locomotor control, rehabilitation, and the early detection of neurological and musculoskeletal disorders. While many datasets capture either biomechanical or neural aspects of gait, publicly available multimodal datasets that integrate brain, muscle, kinematic, and ground reaction force recordings remain scarce. This limitation restricts advances in modeling neuromechanical interactions and the development of machine learning approaches for gait classification and rehabilitation technologies. To address this gap, we provide a comprehensive dataset of treadmill walking from 59 healthy adults with a mean age of 24 ± 5 years, representing both sexes and different body mass index categories. Participants walked at three controlled speeds (0.5, 0.75, and 1.0 m/s), with synchronized recordings from scalp electroencephalography, surface electromyography of 12 lower-limb muscles, inertial sensors capturing kinematics, and bilateral force plates measuring three-dimensional forces, moments, and center of pressure. The dataset enables investigations into brain-body interactions, speed-dependent adaptations, and neuromechanical variability, while supporting the benchmarking of computational models for gait analysis.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":6931764114,"compressed_storage_size":4840062385,"resource_type":"Database","access_policy":"Open","topics":["electroencephalography","biomechanics","neuroscience","gait analysis","force plate","kinematics","inertial measurement unit","electromyography"],"source_url":"https://physionet.org/content/multimodal-gait-dataset/1.0.0/"},{"public_project_uuid":"f14d557f-b1ec-423a-96f4-a6df70485c96","slug":"echonext","version":"1.1.1","core_doi":"10.13026/2j37-ba56","version_doi":"10.13026/7cfw-d091","is_latest_version":true,"short_description":"EchoNext is a curated dataset of electrocardiograms (ECGs) paired with echocardiogram-confirmed structural heart disease labels, designed to support the development and validation of machine learning models.","publish_date":"2026-04-30","title":"EchoNext: A Dataset for Detecting Echocardiogram-Confirmed Structural Heart Disease from ECGs","abstract":"This dataset contains a de-identified collection of 100,000 12-lead electrocardiograms (ECGs) with paired structural heart disease (SHD) labels derived from echocardiography, collected at Columbia University Irving Medical Center. Each ECG is provided with raw waveform data sampled at 250 Hz across all 12 leads, along with accompanying demographic and ECG-specific tabular metadata, including age, sex, heart rate, PR interval, QRS duration, and corrected QT interval. Each ECG is annotated with a binary label indicating the presence or absence of structural heart disease based on echocardiographic findings, including reduced left ventricular ejection fraction, increased ventricular wall thickness, significant valvular disease, right ventricular dysfunction, pulmonary hypertension, or pericardial effusion.
\nThis dataset was developed as part of the creation of the Columbia Mini-Model, a lightweight deep learning model for SHD detection from ECGs. The dataset represents a simplified, focused subset of the larger EchoNext training population and was used to evaluate model performance in resource-constrained settings or smaller-scale deployment environments. It is being released to promote transparency and reproducibility, support further research in cardiovascular AI, and enable benchmarking of lightweight ECG-based screening models for structural heart disease.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":24023103425,"compressed_storage_size":6486764057,"resource_type":"Database","access_policy":"Restricted","topics":["digital health","structural heart disease","electrocardiogram","health equity","artificial intelligence","clinical decision support","ecg","heart failure","transthoracic echocardiogram","ai model deployment","valvular heart disease","machine learning","cardiovascular screening","ai in healthcare","left ventricular dysfunction","deep learning","population health","aortic stenosis"],"source_url":"https://physionet.org/content/echonext/1.1.1/"},{"public_project_uuid":"05af9e75-66b9-4726-8012-063a34432ae9","slug":"b2ai-voice-pediatric","version":"1.1.0","core_doi":"10.13026/mf9s-5r03","version_doi":"10.13026/h995-bt35","is_latest_version":true,"short_description":"A dataset of questionnaire responses, spectrograms, and other information for pediatric participants collected for the Bridge2AI voice as a biomarker of health project.","publish_date":"2026-05-01","title":"Bridge2AI-Voice Pediatric Dataset","abstract":"The human voice contains complex acoustic markers which have been linked to important health conditions including dementia, mood disorders, and cancer. When viewed as a biomarker, voice is a promising characteristic to measure as it is simple to collect, cost-effective, and has broad clinical utility. Recent advances in artificial intelligence have provided techniques to extract previously unknown prognostically useful information from dense data elements such as images. The Bridge2AI-Voice project seeks to create an ethically sourced flagship dataset to enable future research in artificial intelligence and support critical insights into the use of voice as a biomarker of health. Here we present Bridge2AI-Voice, a comprehensive collection of data derived from voice recordings with corresponding clinical information.
\nBridge2AI-Voice Pediatric Dataset v1.1.0 contains derived audio features for 23,533 recordings collected from 300 participants aged 2-18. The release contains data considered low risk, including derivations such as spectrograms but not the original voice recordings. Detailed demographic, clinical, and validated questionnaire data are also made available.
","license":{"name":"Bridge2AI Voice Registered Access License"},"dua":{"name":"Bridge2AI Voice Registered Access Agreement"},"main_storage_size":2940325710,"compressed_storage_size":2897038076,"resource_type":"Database","access_policy":"Credentialed","topics":["voice","health","pediatric","bridge2ai","biomarkers"],"source_url":"https://physionet.org/content/b2ai-voice-pediatric/1.1.0/"},{"public_project_uuid":"51e5c144-c721-4daf-a951-e3bb14155867","slug":"b2ai-voice","version":"3.1.0","core_doi":"10.13026/37yb-1t42","version_doi":"10.13026/8xbn-nq66","is_latest_version":true,"short_description":"A dataset of features from voice recordings and metadata to enable the development, benchmarking, and validation of clinically applicable machine-learning models for diagnosing a wide range of health conditions.","publish_date":"2026-05-01","title":"Bridge2AI-Voice: An ethically-sourced, diverse voice dataset linked to health information","abstract":"The human voice contains complex acoustic markers which have been linked to important health conditions including dementia, mood disorders, and cancer. When viewed as a biomarker, voice is a promising characteristic to measure as it is simple to collect, cost-effective, and has broad clinical utility. Recent advances in artificial intelligence have provided techniques to extract previously unknown prognostically useful information from dense data elements such as images. The Bridge2AI-Voice project seeks to create an ethically sourced flagship dataset to enable future research in artificial intelligence and support critical insights into the use of voice as a biomarker of health. Here we present Bridge2AI-Voice, a comprehensive collection of data derived from voice recordings with corresponding clinical information.
\nBridge2AI-Voice v3.1 contains data for 833 participants across five sites in North America. Participants were selected based on known conditions which manifest within the voice waveform including voice disorders, neurological disorders, mood disorders, and respiratory disorders. The release contains data considered low risk, including derivations such as spectrograms but not the original voice recordings. Detailed demographic, clinical, and validated questionnaire data are also made available.
","license":{"name":"Bridge2AI Voice Registered Access License"},"dua":{"name":"Bridge2AI Voice Registered Access Agreement"},"main_storage_size":13926219167,"compressed_storage_size":13853433651,"resource_type":"Database","access_policy":"Credentialed","topics":["voice","health","bridge2ai","biomarkers"],"source_url":"https://physionet.org/content/b2ai-voice/3.1.0/"},{"public_project_uuid":"514b6375-ac17-4721-8ece-be6de8cedfc2","slug":"bidsleep-dataset","version":"1.0.0","core_doi":"10.13026/hace-p847","version_doi":"10.13026/a0sy-7t69","is_latest_version":false,"short_description":"This dataset contains multi-night recordings of instantaneous heart rate (IHR) and 3-axis accelerometry from the Apple Watch, paired with EEG-based sleep stage labels from the Dreem 2 headband.","publish_date":"2026-05-12","title":"A Multi-Night Instantaneous Heart Rate and Accelerometry Dataset with EEG Sleep Stage Labels","abstract":"We collected sleep data from 47 healthy adult volunteers with no history of sleep disorders, recruited from the local community through study advertisement flyers. The study protocol was approved by the University of Massachusetts Lowell Institutional Review Board. Participants were selected without regard for gender or ethnicity and provided written informed consent. Each participant simultaneously wore a Dreem 2 Headband and an Apple Watch for up to seven consecutive nights, yielding a total of 253 nights of data.
\nThe dataset contains multi-night recordings of instantaneous heart rate and 3-axis accelerometry from the Apple Watch, along with sleep stage labels derived from Dreem 2 EEG and scored according to the AASM standard (Wake, N1, N2, N3, REM). This resource supports the development and validation of wearable-based sleep staging methods, multimodal modeling of heart rate and motion signals, and investigations that benefit from multi-night sleep data.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":29983644385,"compressed_storage_size":6354227911,"resource_type":"Database","access_policy":"Open","topics":["accelerometry","instantaneous heart rate","wearable devices","sleep staging","apple watch"],"source_url":"https://physionet.org/content/bidsleep-dataset/1.0.0/"},{"public_project_uuid":"61a54e2f-c633-47b0-affa-1ab08d392522","slug":"mimic-br","version":"1.0.0","core_doi":"10.13026/x1gx-7j59","version_doi":"10.13026/0vk7-vw29","is_latest_version":false,"short_description":"Medical Information Mart for Intensive Care Brazil (MIMIC-BR) is a Brazilian dataset of hospital and ICU adult patients anonymized data. It includes 31,789 admissions to the Einstein Hospital Israelita during a period of 3 years in the last 10 years","publish_date":"2026-05-21","title":"Medical Information Mart for Intensive Care Brazil (MIMIC-BR): a Brazilian Dataset of Anonymized Hospital and ICU Clinical Data","abstract":"Medical Information Mart for Intensive Care Brazil (MIMIC-BR) is a database of anonymized clinical data from ICU and hospital admissions in Brazil. The dataset format and intended use matches the original large, freely available database comprising de-identified health-related data from patients who were admitted to the critical care units of the Beth Israel Deaconess Medical Center – the original Medical Information Mart for Intensive Care. The MIMIC-BR includes 30,599 adult patients admitted to Einstein Hospital Israelita during a defined retrospective period. The time period consists of three consecutive years within the interval between 2015 and 2025. The exact years are not specified in order to further protect patient privacy. It contains demographic information, diagnoses, procedures, laboratory tests, medications, and vital signs.
\nData are structured using the OMOP Common Data Model (OMOP-CDM) and anonymized in accordance with the Brazilian General Data Protection Law (LGPD), aligned with the current Digital Health Strategy for Brazil 2020-2028, and international best practices to ensure privacy and security. This resource aims to reduce barriers to research, ensure greater representation of Brazilian and Latin American populations in global studies, foster interoperability within the OMOP/OHDSI ecosystem, and promote open science, education, and innovation based on real-world clinical data.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":830207405,"compressed_storage_size":131401348,"resource_type":"Database","access_policy":"Contributor Review","topics":["machine learning","intensive care unit","dataset","critical care","tertiary heatlhcare","data anonymization","inpatients","artificial intelligence"],"source_url":"https://physionet.org/content/mimic-br/1.0.0/"},{"public_project_uuid":"3760ac8b-8e95-48bd-ba77-5f88bbcaa984","slug":"dreamt","version":"2.2.0","core_doi":"10.13026/dztc-dv77","version_doi":"10.13026/3f7y-2d80","is_latest_version":true,"short_description":"We present high resolution wearable device multichannel data along with clinical labeled and recorded sleep stage and polysomnography (PSG) data from 100 sleep abnormal patients with sleep apnea.","publish_date":"2026-06-02","title":"DREAMT: Dataset for Real-time sleep stage EstimAtion using Multisensor wearable Technology","abstract":"Sleep is an intrinsic part of human life, and recent advancements in wearable technology and machine learning have promised continuous and non-invasive methods of tracking sleep health and patterns, providing an important facet to a more holistic understanding of well-being. However, it is still challenging to achieve consistent and reliable real-time estimates of sleep stages using only smartwatches. This is especially true for individuals with irregular sleep patterns or sleep disorders. A major contributing factor is the distinct lack of publicly accessible, large-scale datasets that allow researchers and engineers to validate their wearable sleep staging algorithms against a population with diverse sleep patterns. Here, we present DREAMT, Dataset for Real-time sleep stage EstimAtion using Multisensor wearable Technology, a new dataset collected from 100 participants, which includes high-resolution signals from a smartwatch, expert sleep technician-annotated sleep stage labels, and clinical metadata related to sleep health and disorders.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":122037471161,"compressed_storage_size":21827380848,"resource_type":"Database","access_policy":"Restricted","topics":["time series classification","sleep disorders","wearable","biomedical"],"source_url":"https://physionet.org/content/dreamt/2.2.0/"},{"public_project_uuid":"be8dcccd-2cff-4a10-832e-0efde0d5458d","slug":"inspire","version":"1.4.2","core_doi":"10.13026/dv8j-5412","version_doi":"10.13026/1eay-yc85","is_latest_version":true,"short_description":"A public dataset that contains information related to surgery, anesthesia, laboratory results, medications, diagnosis, and outcomes from 50% of the patients who received surgery at Seoul National University Hospital between 2011 and 2020.","publish_date":"2026-06-09","title":"INSPIRE, a publicly available research dataset for perioperative medicine","abstract":"We present the INSPIRE dataset, a publicly available research dataset in perioperative medicine, which includes approximately 130,000 cases (50% of all surgical cases) who underwent anesthesia for surgery at an academic institution in South Korea between 2011 and 2020. This comprehensive dataset includes patient characteristics such as age, sex, American Society of Anesthesiologists physical status classification, diagnosis, surgical procedure code, department, and type of anesthesia. It also includes vital signs in the operating theatre, general wards, and intensive care units (ICUs), laboratory results from six months before admission to six months after discharge, and medication during hospitalization. Complications include total hospital and ICU length of stay and in-hospital death. We hope this dataset will inspire collaborative research and development in perioperative medicine and serve as a reproducible external validation dataset to improve surgical outcomes.
","license":{"name":"Korea Credentialed Health Data License-1-0-0"},"dua":{"name":"Korea Credentialed Health Data Agreement-1-0-0"},"main_storage_size":537999671,"compressed_storage_size":530244451,"resource_type":"Database","access_policy":"Credentialed","topics":["perioperative medicine","multi-center","surgery","open dataset"],"source_url":"https://physionet.org/content/inspire/1.4.2/"},{"public_project_uuid":"b21d34da-9d46-49f4-80f0-13ebeb118567","slug":"insulin4rl","version":"1.0.0","core_doi":"10.13026/sfvn-y974","version_doi":"10.13026/swen-q904","is_latest_version":true,"short_description":"Openly available research dataset intended for offline reinforcement learning (ORL) using natively irregular healthcare data. The dataset is intended to encourage further research into ORL methods using naturally sporadic decision intervals.","publish_date":"2026-06-16","title":"Insulin4RL: Real-Time Insulin Infusions For Offline Reinforcement Learning","abstract":"Offline reinforcement learning (RL) holds promise for improving healthcare decision-making. However, most existing approaches rely on discretising patient trajectories into strictly regular time intervals. This is done to improve compatibility with neural networks and standard Markov Decision Process (MDP) frameworks, but in doing so, it introduces artefacts of missingness and risks out-of-distribution errors at deployment by training on fictitious representations of the data. Additionally, recent research suggests that retrospective evaluation using binned data can lead to significantly biased estimates of model performance.
\nWe present Insulin4RL, the first freely available dataset designed to encourage RL research on healthcare data with naturally irregular time intervals – both for inputs and labelled decisions. Each label corresponds to a decision regarding the insulin infusion rate, inferred based on proximity to recent blood glucose measurements. This yields over 375,000 labelled decisions from 12,214 patients in the Medical Information Mart for Intensive Care (MIMIC-IV). Clinical rules are applied to robustly identify clinician decisions and subsequent patient states, ensuring immediate compatibility with standard RL methods. Because these decisions occur at irregular intervals, the resulting trajectories can be treated as a semi-Markov Decision Process (SMDP).
\nThe associated state information consists of a sequence of retrospective medical events - including laboratory results, nutritional data, and medication changes - represented as tuples capturing feature, value, and time data. This event-based representation allows direct processing of irregularly sampled data using embedding networks, eliminating the need for discretisation, imputation, or artificially regularised time steps. We hope this approach will encourage research in healthcare RL using data that aligns more closely with real-world practice, using inpatient glycaemic control as a benchmarked example.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":975660402,"compressed_storage_size":963982914,"resource_type":"Database","access_policy":"Credentialed","topics":["insulin","intensive care","semi-markov decision process","diabetes","blood glucose","offline reinforcement learning","machine learning"],"source_url":"https://physionet.org/content/insulin4rl/1.0.0/"},{"public_project_uuid":"e2288d9f-e0f5-4839-9c45-830964978d4a","slug":"argo","version":"1.0.0","core_doi":"10.13026/afxd-9424","version_doi":"10.13026/8gh2-e660","is_latest_version":true,"short_description":"ARGO comprises post-ischemic ventricular tachycardia intracardiac EGMs, 12-lead surface ECGs, and electroanatomical mapping data. It offers a validated ground truth for EGM annotation and delineation by three clinical experts and their consensus.","publish_date":"2026-07-15","title":"Annotated dataset of post-ischemic ventricular tachycardia electrograms (ARGO)","abstract":"The analysis and identification of abnormal ventricular potentials (AVPs) in intracardiac electrograms (EGMs) are paramount in radiofrequency catheter ablation (RFCA) of post-ischemic ventricular tachycardia (VT). The ARGO dataset represents the first open-access dataset of post-ischemic VT EGMs featuring invasive and noninvasive electrophysiological recordings, as well as electroanatomical (EA) map data, including data annotation by three expert electrophysiologists and their final consensus. The dataset includes unipolar and bipolar EGMs, 12-lead surface ECGs, and EA maps, for a total of 1962 entries collected from nine post-ischemic VT patients using the CARTO® 3 mapping system. Every entry encompasses annotations of the bipolar EGMs in three different classes, i.e., Physiological, AVP, or Unknown, as well as the delineation of the onset and end of the pathological activation in the AVPs. The dataset also incorporates detailed information about the spatial localization of ablations performed during the RFCA procedures.
","license":{"name":"Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":396786509,"compressed_storage_size":191810852,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/argo/1.0.0/"},{"public_project_uuid":"ada51cc1-56d4-46a8-bacd-f8e997719292","slug":"brazilian-ophthalmological","version":"1.0.2","core_doi":"10.13026/z9zv-g542","version_doi":"10.13026/mysn-8b26","is_latest_version":true,"short_description":"This is the first Brazilian Multilabel Ophthalmological Dataset with demographic information and retinal photos labeled images according to anatomical parameters, quality control, and presumed diagnosis.","publish_date":"2026-07-27","title":"A Brazilian Multilabel Ophthalmological Dataset (BRSET)","abstract":"The Brazilian Multilabel Ophthalmological Dataset (BRSET) is a multi-labeled ophthalmological dataset designed to improve scientific community development and validate machine learning models. In ophthalmology, ancillary exams support medical decisions and can be used to develop algorithms; however, the availability and representativeness of ophthalmological datasets are limited. This dataset consists of 16,266 images from 8,524 Brazilian patients. Demographics, macula, optic disc, and vessels anatomical parameters, focus, illumination, image field, and artifacts as quality control, and multi-labels are included alongside color fundus retinal photos. This dataset enables computer vision models to predict demographic characteristics and multi-label disease classification using retinal fundus photos.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":16711707281,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Credentialed","topics":["dataset","ophthalmology","retina"],"source_url":"https://physionet.org/content/brazilian-ophthalmological/1.0.2/"},{"public_project_uuid":"c012afc5-2471-4baa-bed6-8bbd45b2097d","slug":"kingston-icu-af","version":"1.0.0","core_doi":"10.13026/gkqx-je58","version_doi":"10.13026/47wc-qk15","is_latest_version":true,"short_description":"A dataset for atrial fibrillation detection research with 596 labelled 10-second ECGs from an ICU in Kingston, Ontario, Canada.","publish_date":"2026-07-29","title":"Kingston ICU AF Dataset","abstract":"This newly published dataset aims to facilitate research in atrial fibrillation (AF) detection in the intensive care unit (ICU). The ICU is a unique environment with noisy signals due to patient movement and prevalent alarms that can lead to alarm fatigue in care providers. The dataset contains 596 labelled 10-second electrocardiograms (ECGs), with 99 atrial fibrillation/flutter samples and 497 sinus rhythm (regardless of the underlying heart rate) samples. These ECGs were collected from bedside monitors in the ICU of the Kingston General Hospital (KGH) in Kingston, Ontario, Canada. Each ECG contains four leads (I, II, III and a V1 equivalent) and was labelled by at least two clinicians.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":23139004,"compressed_storage_size":7419337,"resource_type":"Database","access_policy":"Restricted","topics":["electrocardiography","atrial fibrillation","intensive care unit"],"source_url":"https://physionet.org/content/kingston-icu-af/1.0.0/"},{"public_project_uuid":"a9b1b181-a53c-4418-aa9c-868e8b1687e3","slug":"minute-level-step-count-nhanes","version":"1.0.2","core_doi":"10.13026/5htm-7515","version_doi":"10.13026/d7sw-f662","is_latest_version":true,"short_description":"Minute level step counts obtained from five step counting algorithms for raw accelerometry data, and minute level Activity Counts, MIMS, wear predictions, and wear flags for all participants who wore accelerometers in NHANES 2011-2014.","publish_date":"2026-08-07","title":"Minute level step counts and physical activity data from the National Health and Nutrition Examination Survey (NHANES) 2011-2014","abstract":"The National Health and Nutrition Examination Survey (NHANES) is a nationally representative study that collects demographic, socioeconomic, dietary, and health-related information from 10,000 Americans annually. In Wave G (2011-2012) and Wave H (2013-2014), participants wore an ActiGraph GT3X+ on the non-dominant wrist for seven consecutive days. Publicly available data include minute-level wear predictions (wake wear, sleep wear, unknown, nonwear), data quality flags, and monitor-independent movement summary units (MIMS). Additionally, raw triaxial accelerometer data recorded at 80 Hz are provided, enabling detailed analyses beyond MIMS.
\nWe applied five step counting algorithms and the ActiGraph Activity Count (AC) algorithm to the raw data to create a dataset of minute-level step counts. The dataset also includes minute level AC, MIMS, wear predictions, and wear flags for all participants who wore accelerometers in NHANES 2011–2014. These data facilitate detailed investigations of accelerometry-measured physical activity patterns in a nationally representative U.S. population. Potential analyses include examination of within-day activity patterns and their association with health outcomes.
","license":{"name":"Creative Commons Zero 1.0 Universal Public Domain Dedication"},"dua":null,"main_storage_size":2865163804,"compressed_storage_size":2865040605,"resource_type":"Database","access_policy":"Open","topics":["nhanes","physical activity","accelerometry","steps"],"source_url":"https://physionet.org/content/minute-level-step-count-nhanes/1.0.2/"},{"public_project_uuid":"38ef8cd3-388a-4e26-9774-a27222a2f34b","slug":"bidmc-metabolomic-masld","version":"1.0.0","core_doi":"10.13026/8qyw-qs62","version_doi":"10.13026/a91p-k288","is_latest_version":true,"short_description":"De-identified serum metabolomics from an Australian cohort of adults with and without biopsy-proven MASLD. Includes 1,618 targeted metabolomic features quantified by UPLC-MS/MS (Metabolon Inc.).","publish_date":"2026-08-10","title":"BIDMC Metabolomic Cohort of Adults with Biopsy-Proven MASLD","abstract":"Metabolic dysfunction-associated steatotic liver disease (MASLD), including its progressive form metabolic dysfunction-associated steatohepatitis (MASH), is the leading cause of chronic liver disease. Accurate delineation of the metabolites and signaling pathways that drive hepatic steatosis, inflammation and fibrosis remains essential to support further development of effective screening and treatment strategies.
\nThis project presents a fully de-identified, high-dimensional serum metabolomics dataset derived from a prospective cross-sectional Gastroenterology–Hepatology cohort. Metabolomic analyses and data processing were conducted at Beth Israel Deaconess Medical Center (BIDMC), a tertiary academic center affiliated with Harvard Medical School. Fasted serum samples were analyzed using ultra-high-performance liquid chromatography coupled with tandem mass spectrometry (UPLC-MS/MS), generating 1,618 targeted metabolomic features per participant.
\nThe dataset comprises 110 adult subjects presenting with abnormal liver function tests at a Gastroenterology–Hepatology clinic in the Sydney West Local Health District. This resource is designed to help validate findings from other studies and support hypothesis generation about metabolites involved in MASLD, as well as latent metabolic subgroups across the disease spectrum. The dataset is provided in CSV format with an accompanying README to facilitate secure, transparent, and reproducible reuse under appropriate research agreements.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":2256531,"compressed_storage_size":1006593,"resource_type":"Database","access_policy":"Contributor Review","topics":[],"source_url":"https://physionet.org/content/bidmc-metabolomic-masld/1.0.0/"},{"public_project_uuid":"6367f504-cd34-4dea-9de2-1f4df6d59372","slug":"mimic-br","version":"1.0.1","core_doi":"10.13026/x1gx-7j59","version_doi":"10.13026/mkg6-8h16","is_latest_version":true,"short_description":"Medical Information Mart for Intensive Care Brazil (MIMIC-BR) is a Brazilian dataset of hospital and ICU adult patients anonymized data. It includes 31,789 admissions to the Einstein Hospital Israelita during a period of 3 years in the last 10 years","publish_date":"2026-08-10","title":"Medical Information Mart for Intensive Care Brazil (MIMIC-BR): a Brazilian Dataset of Anonymized Hospital and ICU Clinical Data","abstract":"Medical Information Mart for Intensive Care Brazil (MIMIC-BR) is a database of anonymized clinical data from ICU and hospital admissions in Brazil. The dataset format and intended use match the original large, freely available database comprising de-identified health-related data from patients who were admitted to the critical care units of the Beth Israel Deaconess Medical Center – the original Medical Information Mart for Intensive Care. The MIMIC-BR includes 28,799 adult patients admitted to Einstein Hospital Israelita during a defined retrospective period. The time period consists of three consecutive years within the interval between 2015 and 2025. The exact years are not specified in order to further protect patient privacy. It contains demographic information, diagnoses, procedures, laboratory tests, medications, and vital signs.
\nData are structured using the OMOP Common Data Model (OMOP-CDM) and anonymized in accordance with the Brazilian General Data Protection Law (LGPD), aligned with the current Digital Health Strategy for Brazil 2020-2028, and international best practices to ensure privacy and security. This resource aims to reduce barriers to research, ensure greater representation of Brazilian and Latin American populations in global studies, foster interoperability within the OMOP/OHDSI ecosystem, and promote open science, education, and innovation based on real-world clinical data.
","license":{"name":"PhysioNet Contributor Review Health Data License 1.5.0"},"dua":{"name":"PhysioNet Contributor Review Health Data Use Agreement 1.5.0"},"main_storage_size":844630945,"compressed_storage_size":132982800,"resource_type":"Database","access_policy":"Contributor Review","topics":["critical care","intensive care unit","dataset","machine learning","artificial intelligence","inpatients","tertiary heatlhcare","data anonymization"],"source_url":"https://physionet.org/content/mimic-br/1.0.1/"},{"public_project_uuid":"77910ca7-6736-43f2-80d4-7b5db4e78cdc","slug":"script-carpediem-dataset","version":"1.9.1","core_doi":"10.13026/1qke-qx33","version_doi":"10.13026/n9xd-6d71","is_latest_version":true,"short_description":"SCRIPT contains clinical data and physician adjudication from patients admitted to an intensive care unit, received mechanical ventilation, and underwent bronchoalveolar lavage sampling because of pneumonia suspicion.","publish_date":"2026-08-14","title":"SCRIPT CarpeDiem Dataset: demographics, outcomes, and per-day clinical parameters for critically ill patients with suspected pneumonia","abstract":"Traditional approaches to analyzing episodes of patient care in the ICU examine features on presentation and outcomes on discharge, collapsing the numerous events that happen during a patient’s stay and ignoring intercurrent ICU complications or interventions. With CarpeDiem, we aimed to examine clinical features on a per-day basis as done during the common practice of daily physician rounds. The Successful Clinical Response in Pneumonia Therapy (SCRIPT) CarpeDiem Dataset features 21,931 patient-hospital-days from 704 patients enrolled in the SCRIPT study between June 2018 and May 2023, all of whom were critically ill patients on mechanical ventilation suspected of having pneumonia who underwent a bronchoalveolar lavage as part of routine clinical care. Each patient has demographics, pneumonia episode information as adjudicated by a panel of critical care physicians, outcomes, and 113 clinical parameters for which data are present including vital signs, laboratory parameters, and mechanical support devices. Data have been deidentified per Health Insurance Portability and Accountability Act (HIPAA) Safe Harbor rules. This dataset combines expert clinician adjudication with per-day granular information to provide a unique tool to examine the clinical courses of critically ill patients with pneumonia. We use this dataset in a manuscript examining the contribution of ventilator-associated pneumonia to patient outcomes (see Usage section for details) and share it for others to work with.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":75492525,"compressed_storage_size":13293117,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/script-carpediem-dataset/1.9.1/"},{"public_project_uuid":"4b537321-38a0-42d5-bf0c-5f981846ebd5","slug":"mimic-iv-echo","version":"1.0.1","core_doi":"10.13026/agyq-zp57","version_doi":"10.13026/307c-mr50","is_latest_version":true,"short_description":"The MIMIC-IV-ECHO module contains structured measurements from over 200,000 echocardiograms and more than 500,000 echocardiogram DICOM files. Patients overlap with those in the MIMIC-IV Clinical Database.","publish_date":"2026-08-25","title":"MIMIC-IV-Echo: Echocardiogram Matched Subset","abstract":"The MIMIC-IV-Echo module contains structured echocardiographic measurements and DICOM files from echocardiography exams.
\nStructured measurements are available from 206,488 echocardiogram studies (including 179,928 transthoracic, 16,389 stress, and 10,171 transesophageal echocardiograms). These are from 91,372 unique patients who are in the MIMIC-IV Clinical Database and who had echocardiograms conducted between 2008–2022. The dataset encompasses a broad range of cardiologist-recorded echocardiographic measurements, including chamber sizes and volumes, systolic and diastolic function, valvular morphology and function, aortic root and great vessel dimensions, and Doppler-derived hemodynamics. Clinical interpretations and additional findings are contained in cardiologist free-text reports, which will be made available under the MIMIC-IV-Note module.
\nThe DICOM dataset contains 524,137 files across 7,228 studies from 4,572 distinct patients. This subset contains transthoracic echocardiograms for patients who appear in the MIMIC-IV Clinical Database and were admitted between 2017–2019. A given study consists of numerous sequences of images, with each sequence representing a particular view of the patient's heart.
\nRecords in MIMIC-IV-Echo are matched to the MIMIC-IV Clinical Database, making it possible to link to information across the MIMIC-IV modules. We have also provided information for linking the DICOM files to the structured measurements and notes where available. This linkage supports multimodal research combining echocardiographic imaging and quantitative measurements with longitudinal clinical data.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":1891653079268,"compressed_storage_size":1686860325027,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimic-iv-echo/1.0.1/"},{"public_project_uuid":"531dd05a-7b30-4a47-8f3f-94d0cbf96e8d","slug":"mimic-ext-cxr-qba","version":"1.0.1","core_doi":"10.13026/6193-he91","version_doi":"10.13026/hne5-7t85","is_latest_version":true,"short_description":"We present a large-scale CXR VQA dataset derived from MIMIC-CXR with 42M QA pairs, featuring hierarchical answers, bounding boxes, and structured tags. We generated QA-pairs using LLM-based extraction from radiology reports and localization models.","publish_date":"2026-08-25","title":"MIMIC-Ext-CXR-QBA: A Structured, Tagged, and Localized Visual Question Answering Dataset with Question-Box-Answer Triplets and Scene Graphs for Chest X-ray Images","abstract":"Visual Question Answering (VQA) enables flexible and context-dependent analysis of medical images, such as chest X-rays (CXRs), by allowing users to pose specific questions and receive nuanced answers. However, existing CXR VQA datasets are typically limited to short and simplistic answer, lack localization information (such as bounding boxes), and provide little structured metadata (e.g., hierarchical answer formats or tags like region and finding annotations). To address these limitations, we introduce MIMIC-Ext-CXR-QBA, a new large-scale CXR VQA dataset derived from MIMIC-CXR, comprising 42 million QA pairs, which provides multi-granular, hierarchical answers composed of full sentences in the style of radiology reports, as well as detailed bounding boxes, and structured tags. Additionally, we provide scene graphs for each study, containing both regions and observation nodes with bounding boxes, tags, and textual descriptions derived from the original radiology reports. We created the scene graphs using LLM-based information extraction, semantic mention mapping, and localization models before generating question-answer pairs based on the extracted information stored in these graphs. Using automatic quality assessments, we have selected 31,230,906 QA pairs intended for pre-training and 7,532,281 of these intended for fine-tuning VQA models, therefore providing, to the best of our knowledge, the most sophisticated and largest VQA dataset for CXRs yet.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":28017838528,"compressed_storage_size":27215038201,"resource_type":"Database","access_policy":"Credentialed","topics":["chest x-rays","localization","vqa","scene graphs"],"source_url":"https://physionet.org/content/mimic-ext-cxr-qba/1.0.1/"},{"public_project_uuid":"d09bc268-7a1c-4503-b2c7-af42980e5a51","slug":"mimic-iii-ext-verifact-bhc","version":"1.1.0","core_doi":"10.13026/5cn9-ga59","version_doi":"10.13026/7cqh-bc36","is_latest_version":true,"short_description":"A clinician-labeled dataset for fact-checking long-form clinical text against patient EHRs. The dataset contains LLM-written and human-written Brief Hospital Course summaries decomposed to atomic claim and sentence propositions with annotations.","publish_date":"2026-08-27","title":"MIMIC-III-Ext-VeriFact-BHC: Labeled Propositions From Brief Hospital Course Summaries for Long-form Clinical Text Evaluation","abstract":"The VeriFact-BHC dataset is designed to verify the factuality of long-form text written about a patient against their own electronic health record. There is increasing interest in using large language models (LLMs) to generate clinical text in patient care applications, yet this text needs to be evaluated for factual errors and hallucinations prior to committing text to a patient’s permanent medical record. Text written about a patient should be internally consistent with information already known about the patient, such as that stored in their medical records.
\nVeriFact-BHC contains long-form Brief Hospital Course (BHC) clinical narratives typically found in a discharge summary that have been decomposed into text proposition statements. From 100 patients in the MIMIC-III Clinical Database v1.4, we consider two types of BHC text: a human-written BHC and a LLM-generated BHC. The original human clinician-written BHC is extracted from the discharge summary note. The LLM-generated BHC is composed by a LLM using the patient’s longitudinal clinical notes from the hospital admission. Each BHC is decomposed in two ways: sentence propositions and atomic claim propositions. The remaining electronic health record (EHR) notes for each patient serves as a patient-specific reference of facts that is used by clinicians and VeriFact to assign labels. A total of 13,070 propositions are annotated by multiple clinicians with a ground truth established via majority voting and manual adjudication. Also provided are labels assigned by the VeriFact artificial intelligence system and labels assessing whether propositions are valid from a first-order logic standpoint. The reference EHR for each patient is provided in both machine-readable and PDF formats.
\nBy offering this dataset, we hope to spur further investigation and creation of computational systems for automatic chart review and patient-specific fact verification. We invite the research community to utilize this dataset to develop better methods to guardrail patient-specific LLM-generated clinical text.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":831929975,"compressed_storage_size":830631578,"resource_type":"Database","access_policy":"Credentialed","topics":["clinical informatics","chart review","formal logic","natural language processing","clinical notes","text embedding","electronic health records","artificial intelligence","brief hospital course","long-form text","llm evaluation","logical atomism","retrieval-augmented generation","text reranking","clinical medicine","atomic claim","fact verification","hybrid retrieval","llm-as-a-judge","large language models"],"source_url":"https://physionet.org/content/mimic-iii-ext-verifact-bhc/1.1.0/"},{"public_project_uuid":"59281383-463a-4705-9ffd-317ce9352db3","slug":"mimic-iii-ext-sbdh","version":"1.0.0","core_doi":"10.13026/6ekf-my57","version_doi":"10.13026/ej3k-2436","is_latest_version":true,"short_description":"An annotated social and behavioral determinants of health dataset derived from clinical notes in the MIMIC-III database.","publish_date":"2026-09-01","title":"MIMIC-III-Ext-SBDH: An annotated social and behavioral determinants of health dataset","abstract":"This study presents MIMIC-III-Ext-SBDH, a manually reviewed and consolidated annotation dataset of social and behavioral determinants of health (SBDH) derived from clinical notes in the MIMIC-III database. We previously developed SBDH-Reader, a large language model (LLM)-powered method, to extract granular SBDH data from clinical notes through prompt engineering. During the iterative process of refining the SBDH-Reader prompt, we identified several instances where the original ground truth labels, derived from two previously annotated datasets, were inaccurate. These inconsistencies were systematically reviewed and corrected through consensus-based manual annotation, resulting in a consolidated and updated corpus. The final dataset includes 7,225 clinical notes from 6,382 patients across six SBDH categories: employment, housing, marital status, and substance use including alcohol, tobacco, and drug use. The release includes note-level annotations, associated metadata, and checksum files for data integrity verification.
\nMIMIC-III-Ext-SBDH is intended to support the development and evaluation of natural language processing and (agentic) LLM methods for extracting SBDH information from unstructured clinical narratives and to facilitate reproducible research on socially informed clinical prediction and phenotyping.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":84077496,"compressed_storage_size":28436474,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimic-iii-ext-sbdh/1.0.0/"},{"public_project_uuid":"5b86a9f9-20ec-4a5e-bd88-ebc7aceb9db4","slug":"chexmask-u","version":"1.0.0","core_doi":"10.13026/9yg7-dm71","version_doi":"10.13026/6vn3-3j51","is_latest_version":true,"short_description":"CheXmask-U provides anatomical segmentation masks of chest X-rays derived from landmarks with node-wise uncertainty estimates, enabling landmark-level and image-level analysis for uncertainty-aware research and robustness evaluation.","publish_date":"2026-09-02","title":"CheXmask-U: Uncertainty Estimation for Landmark-Based Anatomical Segmentation Masks in Chest X-ray Images","abstract":"We present an extended version of the CheXmask dataset, augmenting the original anatomical segmentation masks of chest X-rays with node-wise uncertainty estimates. The dataset retains the comprehensive coverage of five public sources—ChestX-ray8, CheXpert, MIMIC-CXR-JPG, Padchest, and VinDr-CXR—while providing uncertainty estimates for each anatomical landmark. CheXmask-U enables both landmark-level and image-level uncertainty assessment, supporting uncertainty-aware segmentation research, robustness evaluation, calibration analysis, and out-of-distribution detection in medical imaging. The dataset includes preprocessed landmark-based segmentation masks for the left lung, right lung, and heart with corresponding node-wise uncertainty estimates, as well as the original individual quality control scores for each of the masks based on Reverse Classification Accuracy (RCA). The obtained uncertainty estimates were evaluated using manual landmark annotations, synthetic corruption experiments, and out-of-distribution analyses, demonstrating meaningful relationships between uncertainty and prediction reliability.
","license":{"name":"Creative Commons Attribution 4.0 International Public License"},"dua":null,"main_storage_size":42836604819,"compressed_storage_size":18052237383,"resource_type":"Database","access_policy":"Open","topics":["uncertainty-aware landmark segmentation"],"source_url":"https://physionet.org/content/chexmask-u/1.0.0/"},{"public_project_uuid":"1eebdb42-f8c1-4c71-92af-0a21e7b4d3da","slug":"mimic-iii-ext-mimic-patient","version":"1.0.0","core_doi":"10.13026/n19w-aw59","version_doi":"10.13026/bzzd-w075","is_latest_version":true,"short_description":"MIMIC-Patient is a patient-level dataset derived from MIMIC-III, providing structured JSON representations of hospital admissions that integrate diagnoses, labs, procedures, and clinical notes to support interactive medical decision-making research.","publish_date":"2026-09-11","title":"MIMIC-III-Ext-MIMIC-Patient: Structured Per-Patient JSON Records for Clinical Question Answering","abstract":"This project provides MIMIC-Patient, a structured, per-patient representation of the MIMIC-III database designed for large language model (LLM)–based clinical question answering. For each of 500 admissions, we reconstruct the patient’s episode into a single JSON file that aggregates demographics, admission information, diagnoses, procedures, medications, vital signs, laboratory results, imaging reports, and key free-text clinical notes. The schema is designed to be fine-grained yet compact, so that LLMs can access clinically relevant information without being overwhelmed by the full raw database.
In addition to the dataset, we provide an answer generation script (answer_agent.py) that implements a patient QA system built on keyword-based routing from natural-language questions to the appropriate JSON fields, followed by LLM-based answer synthesis. This enables standardized benchmarking of LLM performance on retrieval, reasoning, and answer stability when operating on structured EHR abstractions.
The resource is intended for researchers working on clinical NLP, medical decision-making agents, and evaluation of LLMs on real-world electronic health records. All data are derived from the de-identified MIMIC-III database and are shared under a credentialed health data license. Scientific analyses and benchmark results using this resource are described in a separate manuscript.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":19433996,"compressed_storage_size":3124350,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/mimic-iii-ext-mimic-patient/1.0.0/"},{"public_project_uuid":"c611111a-5bdf-4f74-bb58-f15f7e12b932","slug":"vtac","version":"1.1","core_doi":"10.13026/z4f3-1f07","version_doi":"10.13026/jg2e-tf67","is_latest_version":true,"short_description":"VTaC is an annotated ventricular tachycardia (VT) arrhythmia alarm database containing over 5,000 waveform recordings with VT alarms from ICU monitors, with each alarm labeled as either true or false by at least two human expert annotators.","publish_date":"2026-09-14","title":"VTaC: A Benchmark Dataset of Ventricular Tachycardia Alarms from ICU Monitors","abstract":"False arrhythmia alarms are a persistent problem in intensive care units despite considerable effort from industrial and academic algorithm developers. Among the various arrhythmias, ventricular tachycardia (VT) is particularly challenging to detect accurately, as achieving both high sensitivity and high positive predictivity has proven difficult. We present an annotated VT alarm database, VTaC (Ventricular Tachycardia annotated alarms from ICUs) consisting of over 5,000 labeled waveform recordings with VT alarms triggered by bedside monitors in the ICUs. Each VT alarm in the dataset was labeled as true or false by at least two independent human expert annotators. The dataset comprises data collected from ICUs in three major US hospitals and includes data from three leading bedside monitor manufacturers, providing a diverse and representative collection of VT alarm waveform data. Each waveform recording comprises at least two electrocardiogram (ECG) leads and one or more pulsatile waveforms, such as photoplethysmogram (PPG or PLETH) and arterial blood pressure (ABP) waveforms. The current release includes over 17,000 waveform recordings, expanding the initial labeled VTaC release with additional waveform data from unlabeled VT alarm events.
","license":{"name":"Creative Commons Attribution-ShareAlike 4.0 International Public License"},"dua":null,"main_storage_size":19279045214,"compressed_storage_size":11695857538,"resource_type":"Database","access_policy":"Open","topics":["ventricular tachycardia","arrhythmia","machine learning","icu false alarms","benchmark dataset"],"source_url":"https://physionet.org/content/vtac/1.1/"},{"public_project_uuid":"b94c9ade-aeb0-474a-9edc-09f11087e617","slug":"ehr-reasoncon","version":"1.0.0","core_doi":"10.13026/t7hc-k382","version_doi":"10.13026/7m91-3010","is_latest_version":true,"short_description":"Reasoning-Intensive Consistency Verification Between Clinical Notes and Structured Tables in Electronic Health Records","publish_date":"2026-09-15","title":"EHR-ReasonCon: Reasoning-Intensive Consistency Verification Between Clinical Notes and Structured Tables in Electronic Health Records","abstract":"Electronic Health Records (EHRs) combine structured data, such as medications and laboratory measurements, with unstructured clinical notes describing patient conditions and clinical events. Although these two data types provide complementary information, discrepancies may arise between them and potentially affect patient safety and clinical decision-making. Existing datasets for note-table consistency verification mainly focus on direct matching of numerical values or individual events. However, real-world clinical documentation often requires medical interpretation, reasoning over relations between events, and analysis of changes over time. To address this limitation, we developed EHR-ReasonCon, a reasoning-intensive dataset for verifying consistency between clinical notes and structured EHR tables. EHR-ReasonCon was constructed from MIMIC-III in collaboration with clinical experts and contains manual annotations of 8,042 entities from 105 clinical notes. Each entity was checked against records from 14 structured tables and labeled as CONSISTENT or INCONSISTENT.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":70548466,"compressed_storage_size":70239427,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/ehr-reasoncon/1.0.0/"},{"public_project_uuid":"87fb8e70-aa00-4283-b1c3-0287fd00555c","slug":"sex-differences-hearing-loss","version":"1.0.0","core_doi":"10.13026/0xbw-wr52","version_doi":"10.13026/76jk-9b36","is_latest_version":true,"short_description":"The purpose of this study was to obtain sex- and age-specific reference level for hearing threshold of normal aging in Korean and to determine if the rate of change in pure-tone hearing thresholds differed by sex.","publish_date":"2026-09-16","title":"Sex differences in age-related hearing loss","abstract":"The purpose of this study was to obtain sex- and age-specific reference levels for hearing threshold of normal aging in Koreans and to determine if the rate of change in pure-tone hearing thresholds differed by sex. Subjects were clients who visited the Health Promotion Center, Daegu Fatima Hospital, Daegu, Korea, from January 2004 to April 2010. Pure-tone audiometric measures were obtained from 16,884 subjects who also completed a hearing questionnaire. Data were excluded for the present analyses if the subjects were foreigners, repeated testers, or did not complete the questionnaire. Also excluded were subjects with a positive history of otorrhea for any ear, usage of intravenous antibiotics, diuretics, and chemotherapeutic agents, head injury, job in a noisy environment, and military service. Any subjects with asymmetric hearing results were also excluded. More than a 15 dB difference at 0.5, 1, or 2 kHz was defined as asymmetric hearing. Subjects younger than 18 years were excluded to ensure completion of puberty among female subjects. Because no men were older than 80 years, women older than 80 years were excluded to match the age range between sexes. Finally, 4,084 subjects, 910 men and 3,174 women, were included. As part of the health check, subjects completed continuous pure tone audiometric testing followed by pulsed pure-tone testing. Hearing threshold levels were obtained at six frequencies (0.25, 0.5, 1, 2, 4, and 8 kHz). This dataset provides reference values for age-related hearing loss in a carefully screened normal aging population.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":189369,"compressed_storage_size":0,"resource_type":"Database","access_policy":"Open","topics":[],"source_url":"https://physionet.org/content/sex-differences-hearing-loss/1.0.0/"},{"public_project_uuid":"8c259bf1-9483-4e17-bf78-36bc69e9e5ee","slug":"inredd-pan924-ext-boneloss-769","version":"1.0.0","core_doi":"10.13026/yxw0-qr55","version_doi":"10.13026/s6wh-nq14","is_latest_version":true,"short_description":"769 panoramic dental radiographs from InReDD-Dataset-PAN924 with 21,064 COCO polygon annotations for periodontal bone-loss assessment: tooth crowns plus alveolar ridge, maxillary and mandibular reference lines.","publish_date":"2026-09-17","title":"InReDD-Dataset-PAN924-Ext-BoneLoss-769","abstract":"InReDD-Dataset-PAN924-Ext-BoneLoss-769 is a restricted-access extension of InReDD-Dataset-PAN924 containing 769 panoramic dental radiographs and 21,047 geometric annotations in a single COCO-compatible JSON file. It provides 18,792 tooth-crown polygons, 769 alveolar-ridge polygons, 760 mandibular reference polygons and 726 maxillary reference polygons, together with per-image sex. The annotations were produced by an expert-in-the-loop workflow in which a senior dental undergraduate delineated cementoenamel-junction and alveolar-crest references under specialist supervision, pre-existing detection and instance-segmentation models generated crown candidates geometrically from those references, two master's students in dentistry divided the manual refinement, and a specialist professor adjudicated uncertain cases and approved the final ground truth. Manual review removed 1,487 of 20,296 candidates, principally crowns on impacted or unerupted teeth, duplicate candidates, severely compromised crowns and tooth-detection failures; a later pre-release audit removed 17 further records. In a blinded re-annotation exercise covering 52 crowns in three radiographs chosen across bone-loss severity, the two annotators identified the same crowns and obtained a mean Dice similarity coefficient of 0.89 and a mean intersection over union of 0.81. Separate downstream YOLO11x-seg benchmarks reached mask mAP@.5:.95 of 70.39% for crowns and 89.63% for the alveolar ridge on a fixed patient-wise test set. The resource supports crown instance segmentation, cementoenamel junction to alveolar crest geometric analysis, and methodological research on periodontal bone-loss assessment.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":1180171119,"compressed_storage_size":1137815862,"resource_type":"Database","access_policy":"Restricted","topics":["alveolar ridge","dentistry","medical imaging","tooth segmentation","maxilla","computer-aided diagnosis","coco format","crown segmentation","periodontitis","panoramic radiograph","mandible","deep learning","instance segmentation","dental radiograph","periodontal bone loss","oral health"],"source_url":"https://physionet.org/content/inredd-pan924-ext-boneloss-769/1.0.0/"},{"public_project_uuid":"cb5b669b-e389-43dd-aedf-c2ccdba6e777","slug":"jsmf-access","version":"1.0.0","core_doi":"10.13026/98qr-t589","version_doi":"10.13026/m2hn-w830","is_latest_version":true,"short_description":"Processed intraoperative frontal EEG data from 646 surgical patients under general anaesthesia. The dataset includes derived EEG measures for studying anaesthetic brain states, emergence trajectories, and postoperative delirium screening.","publish_date":"2026-09-21","title":"JSMF ACCESS intraoperative derived EEG data set","abstract":"This resource shares processed frontal electroencephalography (EEG) data collected from adult patients undergoing non-emergency, non-cardiac surgery with general anaesthesia at four hospitals. The dataset was generated to study neurophysiological signatures of emergence from anaesthesia and their association with delirium in the postanaesthesia care unit (PACU). The cohort includes data for 646 patients, and delirium was assessed in 626 patients using the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU) after surgery. EEG was recorded intraoperatively.
\nSummarized EEG parameters are included in this release, while full raw EEG (EDF format) is available upon request.
\nThe dataset is intended for research on perioperative brain-state dynamics and prediction of postoperative cognitive complications.
","license":{"name":"PhysioNet Restricted Health Data License 1.5.0"},"dua":{"name":"PhysioNet Restricted Health Data Use Agreement 1.5.0"},"main_storage_size":741963,"compressed_storage_size":226727,"resource_type":"Database","access_policy":"Restricted","topics":["anesthesia eeg delirium surgery burstsupression"],"source_url":"https://physionet.org/content/jsmf-access/1.0.0/"},{"public_project_uuid":"0e3be7d2-21df-4a84-a97b-a20401cc14e7","slug":"mimic-iv-ext-entitycoding","version":"1.0.0","core_doi":"10.13026/h9tg-1h61","version_doi":"10.13026/58h9-pm85","is_latest_version":true,"short_description":"Clinical Named Entity Recognition and Assertion Classification annotations for 400 MIMIC-IV discharge summaries, designed to support ICD-10-CM/PCS coding research.","publish_date":"2026-09-23","title":"MIMIC-IV-Ext-EntityCoding: Clinical Named Entity Recognition and Assertion Classification Dataset for ICD Coding","abstract":"This dataset provides clinical Named Entity Recognition (NER) and Assertion Classification (AC) annotations for 400 discharge summaries from MIMIC-IV-Note (v2.2), designed to support research on computer-assisted ICD coding and related clinical natural language processing tasks. It comprises 85,328 entity annotations across six categories (Normal Finding, Abnormal Finding, Disorder, Procedure, Health Context, and Medication) and 10,922 assertion labels across five categories (Present, Absent, Possible, Hypothetical, and Not associated with patient). In contrast to prior clinical NER resources, the schema was developed with the ICD-10-CM/PCS coding framework in mind, capturing coding-relevant modifiers such as acuity, laterality, and anatomical site within entity spans. The resource supports entity-centric clinical coding pipelines that use annotated entities as compact downstream inputs, and more broadly enables clinical NER and AC research on MIMIC-IV discharge summaries.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":10881038,"compressed_storage_size":2893961,"resource_type":"Database","access_policy":"Credentialed","topics":["clinical nlp","assertion classification","named entity recognition","clinical annotation","clinical coding","discharge summary"],"source_url":"https://physionet.org/content/mimic-iv-ext-entitycoding/1.0.0/"},{"public_project_uuid":"d0242e75-17a6-42a3-9078-71e6dd5226da","slug":"bidsleep-dataset","version":"1.0.1","core_doi":"10.13026/hace-p847","version_doi":"10.13026/rees-1092","is_latest_version":true,"short_description":"This dataset contains multi-night recordings of instantaneous heart rate (IHR) and 3-axis accelerometry from the Apple Watch, paired with EEG-based sleep stage labels from the Dreem 2 headband.","publish_date":"2026-09-25","title":"A Multi-Night Instantaneous Heart Rate and Accelerometry Dataset with EEG Sleep Stage Labels","abstract":"We collected sleep data from 47 healthy adult volunteers with no history of sleep disorders, recruited from the local community through study advertisement flyers. The study protocol was approved by the University of Massachusetts Lowell Institutional Review Board. Participants were selected without regard for gender or ethnicity and provided written informed consent. Each participant simultaneously wore a Dreem 2 Headband and an Apple Watch for up to seven consecutive nights, yielding a total of 253 nights of data.
\nThe dataset contains multi-night recordings of instantaneous heart rate and 3-axis accelerometry from the Apple Watch, along with sleep stage labels derived from Dreem 2 EEG and scored according to the AASM standard (Wake, N1, N2, N3, REM). This resource supports the development and validation of wearable-based sleep staging methods, multimodal modeling of heart rate and motion signals, and investigations that benefit from multi-night sleep data.
","license":{"name":"Open Data Commons Attribution License v1.0"},"dua":null,"main_storage_size":29983644385,"compressed_storage_size":6354227911,"resource_type":"Database","access_policy":"Open","topics":["apple watch","wearable devices","sleep staging","accelerometry","instantaneous heart rate"],"source_url":"https://physionet.org/content/bidsleep-dataset/1.0.1/"},{"public_project_uuid":"30eb8abf-7d63-4422-917c-1291b9991f3b","slug":"cardiowave-portable","version":"1.0.0","core_doi":"10.13026/0q3c-8h90","version_doi":"10.13026/yasp-vq73","is_latest_version":true,"short_description":"A controlled-access portable ECG and PPG waveform database with cuff-based blood pressure labels and benchmark protocols for cuffless blood pressure estimation.","publish_date":"2026-09-30","title":"CardioWave-Portable: a portable ECG and PPG dataset for cuffless blood pressure estimation","abstract":"CardioWave-Portable is a controlled-access dataset of synchronized single-lead electrocardiography (ECG) and photoplethysmography (PPG) signals acquired with a portable card-type device and paired with cuff-based systolic and diastolic blood pressure measurements. The release contains 3,849 fixed-length 24-second waveform records from 977 adult participants, together with record-level blood-pressure labels, signal-quality annotations, and questionnaire-derived participant metadata.
\nThe waveforms are provided as standard WFDB records, with record-level and participant-level metadata provided as CSV files. The dataset is intended to support research on cuffless blood-pressure estimation, physiological signal processing, signal-quality-aware modelling, participant-level evaluation, and reproducible benchmarking. It is a research dataset and does not establish clinical equivalence to validated cuff-based devices.
\nBecause the dataset contains human physiological signals and health-related indirect identifiers, the complete data files are intended for credentialed controlled access under the applicable PhysioNet data-use agreement.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":386348214,"compressed_storage_size":109148639,"resource_type":"Database","access_policy":"Credentialed","topics":[],"source_url":"https://physionet.org/content/cardiowave-portable/1.0.0/"},{"public_project_uuid":"c327d705-6321-437b-bcb4-ca517af00831","slug":"mimic-cxr-ext-mimic-cxr-db","version":"1.0","core_doi":"10.13026/pmv9-jd43","version_doi":"10.13026/gy0s-e371","is_latest_version":true,"short_description":"MIMIC-CXR-DB is a de-identified database linking EHR data with chest X-ray metadata and structured clinical findings. It supports multimodal research by integrating imaging, laboratory, diagnostic, and ontology-linked information.","publish_date":"2026-10-01","title":"MIMIC-CXR-Ext-MIMIC-CXR-DB: Linking MIMIC-IV Chest Radiographs with Rich and Harmonized Clinical Context","abstract":"MIMIC-CXR-DB is a relational clinical database comprising 146,333 chest radiographs linked to hospital admissions, designed to support research at the intersection of electronic health records (EHRs) and chest radiography. The database integrates hospital admissions data, demographic characteristics, diagnoses, laboratory measurements, radiology metadata, radiographic findings, and discharge documentation. It is derived primarily from MIMIC-IV and MIMIC-CXR-JPG, enabling admission-level and imaging-based multimodal analyses. Radiographic findings are derived from chest X-ray reports in the MIMIC-CXR-JPG dataset and include standardized indicators of common thoracic conditions such as atelectasis, cardiomegaly, consolidation, edema, pneumonia, pneumothorax, and pleural abnormalities, alongside detailed DICOM-derived imaging metadata. Laboratory data are normalized using dictionary tables to enable consistent interpretation of test results across admissions. Clinical findings from free-text radiology and discharge reports are extracted as triples using Named Entity Recognition and Ontology-Based Entity Linking and represented in two dedicated tables to improve the discoverability of clinical insights. All data are de-identified and derived from the MIMIC ecosystem under established privacy and data-access requirements. By providing admission-level linkage between clinical variables and imaging studies, MIMIC-CXR-DB enables reproducible research in clinical epidemiology, radiology, and machine learning, particularly for the development and evaluation of models that combine electronic health records with medical imaging.
","license":{"name":"PhysioNet Credentialed Health Data License 1.5.0"},"dua":{"name":"PhysioNet Credentialed Health Data Use Agreement 1.5.0"},"main_storage_size":493247493,"compressed_storage_size":72763359,"resource_type":"Database","access_policy":"Credentialed","topics":["clinical findings","mimic-iv","multimodal research","ehr integration","chest x-ray imaging","ontology-based linking"],"source_url":"https://physionet.org/content/mimic-cxr-ext-mimic-cxr-db/1.0/"}]