HeartBeatAI: An Interpretable and Robust Deep Learning Framework for Multi-Label ECG Arrhythmia Detection
HeartBeatAI is a new deep learning framework designed for multi-label ECG arrhythmia detection. It addresses challenges such as class imbalance and generalization gaps in clinical settings. The framework demonstrates high performance in controlled conditions but struggles with rare anomaly detection in cross-institutional applications.
- ▪HeartBeatAI combines domain generalization, multi-scale feature aggregation, and clinical explainability for ECG classification.
- ▪The framework employs techniques like MixStyle regularization and Label Smoothing to mitigate domain shift.
- ▪It achieved a 98% Macro F1-score in intra-source evaluations but showed significant degradation in detecting rare anomalies during LODO evaluations.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Artificial Intelligence arXiv:2605.24588 (cs) [Submitted on 23 May 2026] Title:HeartBeatAI: An Interpretable and Robust Deep Learning Framework for Multi-Label ECG Arrhythmia Detection Authors:Shubham Gupta, Nikhil Panwar, Partha Pratim Roy View a PDF of the paper titled HeartBeatAI: An Interpretable and Robust Deep Learning Framework for Multi-Label ECG Arrhythmia Detection, by Shubham Gupta and 2 other authors View PDF HTML (experimental) Abstract:While Deep Learning (DL) enhances automated electrocardiogram (ECG) analysis, clinical deployment is hindered by class imbalance and the generalization gap.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
