HeartBeatAI: An Interpretable and Robust Deep Learning Framework for Multi-Label ECG Arrhythmia Detection
arXiv:2605.24588v1 Announce Type: cross Abstract: While Deep Learning (DL) enhances automated electrocardiogram (ECG) analysis, clinical deployment is hindered by class imbalance and the generalization gap. This paper presents HeartBeatAI, a deep learning framework combining domain generalization, multi-scale feature aggregation, and clinical explainability for robust 12-lead ECG classification. Moving beyond image-based paradigms, HeartBeatAI integrates a Squeeze-and-Excitation (SE) ResNet to isolate diagnostic leads alongside a Multi-Layer Concentration Pipeline to capture macro-rhythm and m
The continuous advancements in deep learning necessitate more robust and interpretable AI solutions for critical applications like medical diagnostics.
This development addresses key hurdles in clinical AI deployment—class imbalance, generalization gap, and explainability—paving the way for wider adoption of AI in healthcare.
The clinical utility of AI in ECG analysis is enhanced by a framework designed for both accuracy and transparency, moving beyond previous image-based limitations.
- · Healthcare providers
- · Medical AI developers
- · Patients with cardiac conditions
- · Deep learning researchers
- · Traditional ECG diagnostic methods
- · AI frameworks lacking interpretability
More accurate and reliable AI-driven arrhythmia detection becomes available in clinical settings.
Increased trust in AI diagnostics could lead to accelerated regulatory approvals and wider integration of AI into other medical fields.
The demonstrated demand for explainable and robust AI from the medical sector could influence development priorities across the entire AI industry.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG