MetaSTH-Sleep: Towards Effective Few-Shot Sleep Stage Classification with Spatial-Temporal Hypergraph Enhanced Meta-Learning

2025-05-22

AI Summary

Addressing three key challenges in few-shot cross-subject sleep staging—severe label scarcity, substantial inter-subject variability, and inadequate modeling of high-order spatiotemporal dependencies in physiological signals—this paper proposes the first framework integrating spatial-temporal hypergraph embedding with model-agnostic meta-learning (MAML). Our approach employs a learnable hypergraph neural network to jointly capture heterogeneous spatial relationships among EEG channels, dynamic temporal dependencies, and multiscale features, augmented by graph attention to enhance sensitivity to discriminative substructures. The framework achieves rapid adaptation to unseen subjects using only 1–5 labeled samples per sleep stage. Evaluated on multiple public benchmarks under cross-subject few-shot settings, it attains an average accuracy improvement of 9.2% over state-of-the-art methods, demonstrating significantly enhanced generalization robustness.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Search and Optimization: Metareasoning and Metaheuristics
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Accurate classification of sleep stages based on bio-signals is fundamental for automatic sleep stage annotation. Traditionally, this task relies on experienced clinicians to manually annotate data, a process that is both time-consuming and labor-intensive. In recent years, deep learning methods have shown promise in automating this task. However, three major challenges remain: (1) deep learning models typically require large-scale labeled datasets, making them less effective in real-world settings where annotated data is limited; (2) significant inter-individual variability in bio-signals often results in inconsistent model performance when applied to new subjects, limiting generalization; and (3) existing approaches often overlook the high-order relationships among bio-signals, failing to simultaneously capture signal heterogeneity and spatial-temporal dependencies. To address these issues, we propose MetaSTH-Sleep, a few-shot sleep stage classification framework based on spatial-temporal hypergraph enhanced meta-learning. Our approach enables rapid adaptation to new subjects using only a few labeled samples, while the hypergraph structure effectively models complex spatial interconnections and temporal dynamics simultaneously in EEG signals. Experimental results demonstrate that MetaSTH-Sleep achieves substantial performance improvements across diverse subjects, offering valuable insights to support clinicians in sleep stage annotation.

Problem

• Addresses limited labeled data in sleep stage classification
• Improves generalization across diverse individual bio-signals
• Models high-order spatial-temporal relationships in EEG signals

Innovation

• Spatial-temporal hypergraph enhances EEG signal modeling
• Meta-learning enables few-shot adaptation to new subjects
• Simultaneously captures signal heterogeneity and temporal dynamics