Physiology-Aware Masked Cross-Modal Reconstruction for Biosignal Representation Learning

TL;DR

This paper introduces xMAE, a biosignal pretraining framework that leverages masked cross-modal reconstruction to encode physiologically meaningful timing structures in multimodal biosignals, improving downstream task performance.

Contribution

xMAE is the first to incorporate temporal structure into multimodal biosignal pretraining, capturing directional dynamics between signals like ECG and PPG.

Findings

Pretraining with xMAE outperforms unimodal and multimodal baselines on 15 of 19 tasks.

xMAE representations generalize across devices, locations, and settings.

ECG-PPG timing structure is reflected in learned PPG representations.

Abstract

Biosignals acquired from different locations on the body often provide temporally ordered views of the same underlying physiological process. However, most existing self supervised learning methods treat these signals as interchangeable views, overlooking the directional temporal dynamics that link them. A canonical example is the relationship between electrocardiography (ECG), which captures the electrical activation initiating each heartbeat, and photoplethysmography (PPG), which records the resulting peripheral pulse delayed by vascular dynamics. To capture this structured relationship, we introduce xMAE, a biosignal pretraining framework that leverages masked cross modal reconstruction across temporally ordered biosignals as a training time constraint to encourage physiologically meaningful timing structure in the learned representations. We show that pretraining with xMAE yields representations that outperform both unimodal and multimodal baselines on 15 of 19 downstream tasks, including cardiovascular outcome prediction, abnormal laboratory test detection, sleep staging, and demographic inference, while generalizing across devices, body locations, and acquisition settings. Further analysis suggests that the ECG PPG timing structure is reflected in the learned PPG representations. More broadly, xMAE demonstrates the effectiveness of incorporating temporal structure into multimodal pretraining when signals observe different stages of a shared underlying process. Code is available at https://github.com/hzhou3/xMAE.