PedSleepMAE: Generative Model for Multimodal Pediatric Sleep Signals

TL;DR

PedSleepMAE is a novel generative model that leverages multimodal pediatric sleep signals to improve sleep scoring, detect sleep disorders, and generate realistic signals for various clinical applications.

Contribution

It introduces the first general-purpose generative model trained on multiple pediatric sleep signal modalities, enhancing analysis and interpretation.

Findings

Comparable sleep scoring performance to supervised models

Effective detection of apnea, hypopnea, EEG arousal, and oxygen desaturation

Generates realistic signals for retrieval, outlier detection, and imputation

Abstract

Pediatric sleep is an important but often overlooked area in health informatics. We present PedSleepMAE, a generative model that fully leverages multimodal pediatric sleep signals including multichannel EEGs, respiratory signals, EOGs and EMG. This masked autoencoder-based model performs comparably to supervised learning models in sleep scoring and in the detection of apnea, hypopnea, EEG arousal and oxygen desaturation. Its embeddings are also shown to capture subtle differences in sleep signals coming from a rare genetic disorder. Furthermore, PedSleepMAE generates realistic signals that can be used for sleep segment retrieval, outlier detection, and missing channel imputation. This is the first general-purpose generative model trained on multiple types of pediatric sleep signals.