Pretraining on Sleep Data Improves non-Sleep Biosignal Tasks

TL;DR

This study explores using sleep biosignals as pretraining data to enhance performance on various non-sleep biosignal tasks, demonstrating consistent improvements across multiple datasets and tasks.

Contribution

It introduces a sleep-based contrastive pretraining approach that transfers effectively to non-sleep EEG and ECG tasks, often surpassing existing models.

Findings

Sleep pretraining improves performance on 8 downstream tasks.

Achieves results comparable to or better than state-of-the-art models.

Effective transfer across multiple EEG and ECG datasets.

Abstract

Sleep foundation models have recently demonstrated strong performance on in-domain polysomnography tasks, including sleep staging, apnea detection, and disease risk prediction. In this work, we investigate whether sleep biosignals can serve as an effective pretraining distribution for learning representations that transfer beyond sleep to adjacent domains. Following sleep foundation models, we perform sleep-only multimodal contrastive pretraining (with a leave-one-out objective) and evaluate transfer to non-sleep EEG and ECG, two well-benchmarked biosignal modalities with heterogeneous datasets and clinically meaningful downstream tasks. Across eight downstream tasks spanning multiple EEG and ECG datasets, sleep pretraining consistently improves performance relative to training from scratch. Moreover, on several tasks, we achieve performance competitive with or surpassing prior specialized state-of-the-art and foundation models.