EXGnet: a single-lead explainable-AI guided multiresolution network with train-only quantitative features for trustworthy ECG arrhythmia classification

Tushar Talukder Showrav, Soyabul Islam Lincoln, Md. Kamrul Hasan

Deep learning has significantly propelled the performance of ECG arrhythmia classification, yet its clinical adoption remains hindered by challenges in interpretability and deploym…