Modeling Day-Long ECG Signals to Predict Heart Failure Risk with Explainable AI
Researchers have developed a deep learning model called DeepHHF that can predict the risk of heart failure within five years using 24-hour electrocardiogram (ECG) data. The model, trained on the Technion-Leumit Holter ECG (TLHE) dataset, achieved an area under the receiver operating characteristic curve of 0.80, outperforming traditional methods. Explainability analysis revealed that DeepHHF focuses on arrhythmias and heart abnormalities, highlighting the potential of AI in non-invasive and accessible heart failure risk prediction. AI
IMPACT This research demonstrates the potential for AI to improve non-invasive and accessible heart failure risk prediction, potentially leading to earlier interventions and better patient outcomes.