Omni-Sleep: A Sleep Foundation Model via Hierarchical Contrastive Learning of CNS--ANS Dynamic

2026-07-04

AI Summary

This work addresses the limitation of existing sleep foundation models, which overlook the hierarchical organization of the central nervous system (CNS) and autonomic nervous system (ANS) underlying multimodal physiological signals, thereby failing to capture their dynamic interplay. To overcome this, the study introduces a topology-constrained hierarchical contrastive learning framework that explicitly incorporates CNS-ANS physiological partitioning as a prior. The framework jointly optimizes three objectives: intra-system consistency, inter-system synchrony, and masked temporal modeling in latent space, to learn coordinated brain-body representations. Pretrained on over 100,000 hours of multicenter polysomnography (PSG) data, the model significantly outperforms current methods in both sleep staging and multi-disease classification tasks, demonstrating superior label efficiency, cross-dataset generalizability, and robustness to missing modalities.

Abstract

Sleep physiology arises from the coordinated dynamics of the central nervous system (CNS) and autonomic nervous system (ANS), as reflected by multimodal polysomnography signals including EEG, EOG, EMG, ECG, and respiration. However, existing sleep foundation models often fuse heterogeneous biosignals in a topology-agnostic manner, overlooking their physiological organization. We introduce Omni-Sleep, a sleep foundation model that uses the CNS/ANS partition as a physiological prior for topology-constrained representation learning. Omni-Sleep learns structured representations through three objectives: intra-system consistency, which captures shared subsystem-level factors within neural and cardio-respiratory signals; inter-system synchronization, which aligns subsystem trajectories to model brain--body dynamics; and latent-space masked temporal modeling, which captures long-horizon sleep dynamics. Pre-trained on over 100,000 hours of multi-center multimodal PSG data, Omni-Sleep is evaluated on sleep staging and multi-disease classification. Across datasets and modality-ablation settings, Omni-Sleep outperforms strong foundation-model baselines, showing improved label efficiency, cross-dataset generalization, and robustness to missing modalities. These results highlight the value of physiological hierarchy for generalizable sleep representation learning. Code is available at https://github.com/AutoBrain-sleep/OmniSleep.

Problem

• sleep foundation model
• central nervous system
• autonomic nervous system
• multimodal polysomnography
• physiological hierarchy

Innovation

• hierarchical contrastive learning
• physiological prior
• CNS-ANS dynamics
• multimodal sleep foundation model
• topology-constrained representation