What does this research mean for the field?

The Harvard-Emory ECG Database is the largest open-access collection of 12-lead ECG recordings, comprising over 10 million recordings from nearly 2 million patients. Novelty: ClaimNovelty.SYNTHESIS. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

The aim is to develop a comprehensive ECG database to enhance machine learning applications in cardiovascular health.

February 27, 2026Open Access

The Harvard-Emory ECG Database

Key Result

The Harvard-Emory ECG Database provides an open-access collection of 11,607,261 clinical 12-lead ECG recordings from 2,167,795 patients, linked to demographic metadata and diagnostic codes.

Key Points

The aim is to develop a comprehensive ECG database to enhance machine learning applications in cardiovascular health.
Compilation of over 10 million ECG recordings from two major hospitals and associated metadata.
Inclusion of demographic information and diagnostic labels generated by analysis software.
Data collection spanning over four decades, recorded in clinical settings.
The dataset contains 10,608,417 ECG recordings linked to over a million patients.
Diverse demographic data accompanies the ECG recordings, supporting broad applicability.
Includes various diagnostic annotations, enabling detailed analysis and machine learning algorithm training.

Structured PICO

Population

2,167,795 patients (1,818,247 from Massachusetts General Hospital and 349,548 from Emory University Hospital) with 11,607,261 12-lead ECG recordings collected between 1980 and 2022 in clinical settings.

The Harvard-Emory ECG Database provides the largest open-access collection of over 11.6 million clinical 12-lead ECGs with linked metadata and diagnostic codes to support machine learning and cardiovascular research.

Limitations

No direct relationship has been established between an individual ECG's automated interpretation, the human correction, and the clinical diagnoses documented in the ICD codes.
Potential limitation in generalizability, as the data originate from a single equipment ecosystem (GE Healthcare, MUSE system), which could introduce vendor-specific bias.
Potential geographical bias, as the data were collected from two academic centers in the United States.
Errors may have been introduced during the mapping of free-text physician statements to standardized 12SL codes.

Abstract

The Harvard-Emory ECG Database (HEEDB) is currently the largest open-access collection of 12-lead electrocardiogram (ECG) recordings, developed through a collaboration between Harvard and Emory University. The database consists of 10,608,417 ECG recordings from 1,818,247 patients from Massachusetts General Hospital (MGH) and 998,844 recordings from 349,548 patients from Emory University Hospital (EUH) collected between 1980 and 2022 in clinical settings as part of routine patient care. The ECGs are 10-second, 12-lead recordings sampled at either 250 or 500 Hz, and stored in WFDB format. Each ECG is linked to demographic metadata (age, sex, race, ethnicity, education), along with deidentified acquisition dates, last visit dates, and death dates, when available. The dataset includes three forms of MarquetteTM 12SL Analysis Software annotations: (1) batch-reprocessed diagnostic labels generated using the latest available 12SL software (version 24); (2) the original 12SL outputs from the time of ECG acquisition; and (3) the corresponding physician overreads. Additionally, the dataset includes associated ICD-9 and ICD-10 codes with the corresponding diagnosis dates. This database represents a large, diverse multi-center collection on which machine learning algorithms can be trained and tested for performance and bias.

Bookmark

View Full Paper