Columbia University Irving Medical Center
Detects 12 categories of echocardiogram-confirmed structural heart disease from 12-lead ECG waveforms combined with demographic and clinical covariates. Uses the same architecture as the original, larger EchoNext model but is trained entirely on the public EchoNext-Mini dataset - 100,000 de-identified ECGs from Columbia University Irving Medical Center released on PhysioNet - making it one of the more fully reproducible models of its kind, with public weights, a Docker image, and inference code.
Architecture
Hybrid
Deep neural network combining a 12-lead ECG waveform encoder with tabular demographic/clinical covariates, with multi-label output over 12 structural-heart-disease categories (same architecture as the original EchoNext model)
Framework
PyTorch
Added to catalog
2026-07-10
EchoNext (PhysioNet) open_in_new
100,000 de-identified 12-lead ECGs from 36,286 adult patients at Columbia University Irving Medical Center, each paired with echocardiogram-derived structural-heart-disease labels from a TTE within 1 year of the ECG.
Composite structural heart disease detection