PhysioNet

PhysioNet is an open-access repository specialized for complex physiological signal data with annotations, as well as software tools and educational research. Topic areas include cardiology, neuroscience, clinical care and machine learning applications. Deposited datasets should ideally be intended for reuse by their research community. Most collections are human data, but non-human are accepted. PhysioNet has two tiers of controlled access options for restricting data, but may not meet JHU sharing policy criteria for higher-risk data. Consult with IRB before depositing.

Repository Website

Repository Scope

Data Collection Policy

Research Areas

Medicine; Anatomy and Physiology; Biophysics; Neurosciences; Medical Informatics; Medical Physics, Biomedical Technology;

Data Types

Biomedical; Image; Code; Machine learning; Synthetic/Simulation; Tabular; Observational;

Data Types Encouraged/Permitted

Physiological signals and waveform data (e.g., ECG, EEG, respiratory signals); medical imaging (MRI, CT); diseases and clinical contexts including cardiovascular; neuroscience formats; biomedical multimodal measures; associated annotations; health data and clinical contexts; derived and synthetic data for machine learning; software and code

Data Types Explicitly Prohibited

Data not for biomedical or physiological research. Non-human data are allowed but most collections are human

Fee for JHU Researchers to Deposit

No

Data Limit

Request approval for deposits above 100GB

Data Access

Data Access Policy

Option for Data Access

Controlled Access; Open Access;

Details on Data Access

  • Open Access: Accessible by all users, with minimal restrictions on reuse.
  • Restricted Access: Accessible by registered users who sign a Data Use Agreement reviewed by PhysioNet.
  • Credentialed Access: Accessible by registered users who complete the credentialing process (CITI Data and Specimens Training, HIPAA Module) and sign a Data Use Agreement. For Restricted and Credentialed access, PhysioNet staff verify credentials and provide broad access to datasets rather than approving specific dataset access requests. Note that this may not meet General JHU IRB Expectations for Sharing Individual Level Research Data for some datasets. Consult with IRB before depositing data.

Sensitive Data

Human Data Accepted

Yes

Level of Deidentification Required

All Open Access data should meet HIPAA statistical determination standards with proper participant consent, with similar criteria for “restricted” and “credentialed” files when feasible, HIPAA Safe Harbor de-identification at minimum. Your de-identification plan may require review and approval by IRB or JHM Data Trust for JHM systems data.

Human Participant Data Sharing Policy

Sensitive Data Policy

Administration

Submission Policy

Required Funder

None listed

Persistent Identifier

DOI

Data Retention Period

Indefinitely

AI LLM Policy

Accepts AI/LLM data and models for biomedical and physiology use

re3data

re3data Keywords:

Resource, PWAVE, aging, biomedical signals, cardiology, eICU, electrophysiology, heart, nervous system, nonlinear dynamics, physiological signals, physiology, sudden death

re3data Repository Contact

https://physionet.org/about/#contact_us

re3data Record