ECG-RAMBA: Zero-Shot ECG Generalization by Morphology-Rhythm Disentanglement and Long-Range Modeling
This addresses the challenge of reliable ECG analysis for clinical deployment and monitoring by improving cross-dataset robustness, though it is incremental as it builds on existing methods like MiniRocket and Mamba with novel integration.
The paper tackled the problem of poor generalization of deep learning models for ECG classification across heterogeneous datasets by proposing ECG-RAMBA, a framework that disentangles morphology and rhythm features and uses long-range modeling, achieving a macro ROC-AUC of approximately 0.85 on the Chapman-Shaoxing dataset and a PR-AUC of 0.708 in zero-shot transfer for atrial fibrillation detection on CPSC-2021.
Deep learning has achieved strong performance for electrocardiogram (ECG) classification within individual datasets, yet dependable generalization across heterogeneous acquisition settings remains a major obstacle to clinical deployment and longitudinal monitoring. A key limitation of many model architectures is the implicit entanglement of morphological waveform patterns and rhythm dynamics, which can promote shortcut learning and amplify sensitivity to distribution shifts. We propose ECG-RAMBA, a framework that separates morphology and rhythm and then re-integrates them through context-aware fusion. ECG-RAMBA combines: (i) deterministic morphological features extracted by MiniRocket, (ii) global rhythm descriptors computed from heart-rate variability (HRV), and (iii) long-range contextual modeling via a bi-directional Mamba backbone. To improve sensitivity to transient abnormalities under windowed inference, we introduce a numerically stable Power Mean pooling operator ($Q=3$) that emphasizes high-evidence segments while avoiding the brittleness of max pooling and the dilution of averaging. We evaluate under a protocol-faithful setting with subject-level cross-validation, a fixed decision threshold, and no test-time adaptation. On the Chapman--Shaoxing dataset, ECG-RAMBA achieves a macro ROC-AUC $\approx 0.85$. In zero-shot transfer, it attains PR-AUC $=0.708$ for atrial fibrillation detection on the external CPSC-2021 dataset, substantially outperforming a comparable raw-signal Mamba baseline, and shows consistent cross-dataset performance on PTB-XL. Ablation studies indicate that deterministic morphology provides a strong foundation, while explicit rhythm modeling and long-range context are critical drivers of cross-domain robustness.