LGFNet: A CTC-Guided Local-Global Fusion Framework for Single-Channel Sleep Staging

2026-07-27

AI Summary

This work addresses the challenges of single-channel sleep staging—particularly the difficulty in modeling long-range temporal dependencies, the ambiguity of N1 stage and transition segments, and distribution shifts across subjects and devices—which are especially pronounced in low-latency wearable settings. To this end, the authors propose LGFNet, a novel framework featuring a local-global fused encoder that jointly captures fine-grained temporal dynamics and long-range sleep architecture. It integrates a CTC-guided attention mechanism with a three-stage decoder incorporating Viterbi smoothing to enhance boundary detection accuracy and physiological plausibility. Evaluated on five public datasets, LGFNet substantially outperforms existing single-channel methods, achieving relative improvements of 1.27% in accuracy, 1.74% in macro-F1, and 1.93% in Cohen’s kappa on Sleep-EDF-78, with notably superior performance on N1 and transitional epochs.

Abstract

Sleep staging remains challenging due to long-range temporal dependencies, ambiguous stage transitions-particularly in N1-and substantial distribution shifts across subjects, sampling rates, and EEG montages. These difficulties are further amplified in single-channel, low-latency scenarios required by wearable and real-world applications. To address these issues, we propose LGFNet, a CTC-guided sequence-to-sequence framework for robust sleep staging. LGFNet introduces a Local-Global Fusion encoder that jointly models fine-grained temporal dynamics and long-range sleep structure, overcoming the limitations of conventional serial hybrid architectures. A CTC-Attention joint training paradigm is adopted to unify temporal alignment with context-dependent modeling, enabling more accurate recognition of stage boundaries and transitions. Furthermore, a three-stage decoding strategy is devised, leveraging CTC-guided decoding and Viterbi-based smoothing to reduce error accumulation and enforce physiological consistency. Extensive cross-dataset evaluations on five public benchmarks demonstrate that LGFNet consistently outperforms state-of-the-art single-channel methods. In particular, on Sleep-EDF-78, LGFNet surpasses DMIN by +1.27% accuracy, +1.74% macro-F1, and +1.93% kappa, with pronounced gains on N1 and transition segments, highlighting its robustness and strong generalization across diverse sampling rates, montages, and recording environments.

Problem

• sleep staging
• single-channel EEG
• temporal dependencies
• stage transitions
• distribution shifts

Innovation

• Local-Global Fusion
• CTC-Attention joint training
• single-channel sleep staging
• sequence-to-sequence framework
• Viterbi-based smoothing