The Harvard-Emory ECG Database
5 hours ago
- #ECG
- #medical-database
- #health-data
- The Harvard-Emory ECG Database (HEEDB) is a large collection of 12-lead ECG recordings from Massachusetts General Brigham and Emory University.
- Version 4.0 includes 10,471,531 ECGs from 1,818,247 unique patients at institution I0001 and 968,680 ECGs from 349,548 patients at institution I0006.
- Data is stored in WFDB and Matlab formats, with metadata, 12SL diagnostic codes, and ICD-9/10 codes included.
- The database is part of the Human Sleep Project (HSP), funded by NHLBI grant R01HL161253.
- ECG recordings are 10 seconds long, sampled at 250 or 500 Hz, and de-identified using the Safe Harbor method.
- Directory structure includes separate folders for diagnoses, ICD codes, metadata, and waveform files for each institution.
- 12SL_diagnoses folder contains diagnostic outputs from GE Healthcare's 12SL software, with mappings to human-readable labels.
- ICD_codes folder includes ICD-9 and ICD-10 codes from EHRs, with descriptions and shifted dates.
- Metadata includes demographic and temporal information such as ECG acquisition time, date of birth, and age-related fields.
- Access is restricted to credentialed users who sign a Data Use Agreement and complete CITI training.
- The study was approved by IRBs at Massachusetts General Hospital and Beth Israel Deaconess Medical Center, with informed consent waived due to retrospective design.
- Dr. Westover has a conflict of interest as a co-founder of Beacon Biosignals; other authors declare no conflicts.