Horace Mann School / Emory University School of Medicine
Real-time deep-learning model that fuses serial 12-lead ECG waveforms with sequential vital signs and routinely available clinical data to predict hospital admission early during emergency department (ED) encounters with cardiac presentations (chest pain, dyspnea, syncope, presyncope). Developed and validated on the public MIMIC-IV, MIMIC-IV-ED, and MIMIC-IV-ECG databases (n=30,421 ED stays with >=1 ECG; n=11,273 with >=2 ECGs), the model improves on baseline tabular (random forest) and ECG-only models by leveraging how a patient's risk evolves with successive ECGs during the visit, addressing a key limitation of single-time-point risk scores.
Architecture
Hybrid
Multimodal deep learning model fusing a 12-lead ECG waveform encoder with sequential vitals and structured clinical variables via a transfer-learning approach, predicting hospital admission from serial ED encounter data
Added to catalog
2026-08-14
800,035 12-lead ECG-report pairs from 161,352 subjects at Beth Israel Deaconess Medical Center.
Real-time predicted probability of hospital admission from serial ECG + vitals + clinical data during an ED encounter