In-Distribution and Out-of-Distribution Self-supervised ECG Representation Learning for Arrhythmia Detection

TL;DR

This study systematically evaluates self-supervised learning methods for ECG arrhythmia detection, demonstrating their effectiveness and strong generalization capabilities across multiple datasets and conditions.

Contribution

It is the first to quantitatively analyze ECG data distributions and compare various SSL methods, highlighting SwAV's superior performance and SSL's robustness in ID and OOD scenarios.

Findings

SSL methods achieve competitive results with supervised approaches.

SSL representations generalize well across different datasets.

SwAV outperforms other SSL methods in ECG representation learning.

Abstract

This paper presents a systematic investigation into the effectiveness of Self-Supervised Learning (SSL) methods for Electrocardiogram (ECG) arrhythmia detection. We begin by conducting a novel analysis of the data distributions on three popular ECG-based arrhythmia datasets: PTB-XL, Chapman, and Ribeiro. To the best of our knowledge, our study is the first to quantitatively explore and characterize these distributions in the area. We then perform a comprehensive set of experiments using different augmentations and parameters to evaluate the effectiveness of various SSL methods, namely SimCRL, BYOL, and SwAV, for ECG representation learning, where we observe the best performance achieved by SwAV. Furthermore, our analysis shows that SSL methods achieve highly competitive results to those achieved by supervised state-of-the-art methods. To further assess the performance of these methods on both In-Distribution (ID) and Out-of-Distribution (OOD) ECG data, we conduct cross-dataset training and testing experiments. Our comprehensive experiments show almost identical results when comparing ID and OOD schemes, indicating that SSL techniques can learn highly effective representations that generalize well across different OOD datasets. This finding can have major implications for ECG-based arrhythmia detection. Lastly, to further analyze our results, we perform detailed per-disease studies on the performance of the SSL methods on the three datasets.