Dynamic Prototype Rehearsal for Continual ECG Arrhythmia Detection
Sana Rahmani, Reetam Chatterjee, Ali Etemad, Javad Hashemi
TL;DR
This work tackles catastrophic forgetting in continual learning for ECG arrhythmia detection by introducing DREAM-CL, a dynamic prototype rehearsal method that builds a memory of representative, difficult samples. The approach clusters training samples by loss-update trajectories, applies a Lambert W-based compression to curb outliers, and replays the hardest representatives across sessions to preserve past knowledge. Evaluated on time-, class-, and lead-incremental settings with Chapman and PTB-XL, DREAM-CL achieves state-of-the-art performance and demonstrates robustness through extensive ablations of clustering, memory size, and the outlier-robust sampling strategy. The results underscore DREAM-CL’s potential to enhance continual ECG analysis in settings with evolving data distributions and limited memory budgets.
Abstract
Continual Learning (CL) methods aim to learn from a sequence of tasks while avoiding the challenge of forgetting previous knowledge. We present DREAM-CL, a novel CL method for ECG arrhythmia detection that introduces dynamic prototype rehearsal memory. DREAM-CL selects representative prototypes by clustering data based on learning behavior during each training session. Within each cluster, we apply a smooth sorting operation that ranks samples by training difficulty, compressing extreme values and removing outliers. The more challenging samples are then chosen as prototypes for the rehearsal memory, ensuring effective knowledge retention across sessions. We evaluate our method on time-incremental, class-incremental, and lead-incremental scenarios using two widely used ECG arrhythmia datasets, Chapman and PTB-XL. The results demonstrate that DREAM-CL outperforms the state-of-the-art in CL for ECG arrhythmia detection. Detailed ablation and sensitivity studies are performed to validate the different design choices of our method.