AUTOMATIC SLEEP STAGE CLASSIFICATION USING MULTI-PHYSIOLOGICAL SIGNALS BASED ON PCBSLEEPNET
Eurasia Journal of Science and Technology · 03.04.2026 · 8 (2) · 1-8
DOI: 10.61784/ejst3140 ↗Kurz gefasst
Fachartikel in der Zeitschrift Eurasia Journal of Science and Technology, online erschienen am 03.04.2026. Verfasst von YuLi Yang und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).
Abstract (Originalsprache)
Automatic sleep staging, a key technology in sleep medicine, faces feature selection challenges due to multi-modal physiological signal heterogeneity. This study establishes a framework to evaluate bioelectrical signals' impacts on deep neural network feature representation, aiming to identify optimal integration paradigms.We propose a multi-branch hybrid neural network with: 1) a three-channel parallel convolutional module incorporating cascaded residual and dilated convolutions for time-frequency feature extraction, and 2) bidirectional long short-term memory (BiLSTM) layers for temporal dependency modeling. We systematically analyzed multi-channel electroencephalogram (EEG), single-channel physiological signals, and their combinations covering both single-channel and multi-channel configurations of EEG, electrooculogram (EOG), and electromyogram (EMG). Results demonstrate that six-channel EEG configurations outperformed single/dual-channel implementations. In the single- and dual-channel EEG settings, the C3-M2 and C4-M1 + C3-M2 channels exhibited superior performance. EEG maintained superior classification capability over EOG/EMG. The tri-modal fusion (EEG+EOG+EMG) achieved peak performance (accuracy: 89.5%, Cohen’s κ: 0.85). This demonstrates EEG's central role in sleep staging, while EOG/EMG synergistically enhance model performance. The findings reveal heterogeneous bioelectric signal interactions in deep neural networks, providing foundations for optimizing signal fusion strategies and model architectures.
Zitieren
APA
Yang, Y., Liu, Y., & Zhang, H. (2026). AUTOMATIC SLEEP STAGE CLASSIFICATION USING MULTI-PHYSIOLOGICAL SIGNALS BASED ON PCBSLEEPNET. Eurasia Journal of Science and Technology, 8(2), 1-8. https://doi.org/10.61784/ejst3140
Vancouver
Yang Y, Liu Y, Zhang H. AUTOMATIC SLEEP STAGE CLASSIFICATION USING MULTI-PHYSIOLOGICAL SIGNALS BASED ON PCBSLEEPNET. Eurasia Journal of Science and Technology. 2026;8(2):1-8. doi:10.61784/ejst3140
BibTeX
@article{Yang2026,
author = {Yang, YuLi and Liu, Ying and Zhang, HaoWei},
title = {AUTOMATIC SLEEP STAGE CLASSIFICATION USING MULTI-PHYSIOLOGICAL SIGNALS BASED ON PCBSLEEPNET},
journal = {Eurasia Journal of Science and Technology},
year = {2026},
volume = {8},
number = {2},
pages = {1--8},
doi = {10.61784/ejst3140}
}
EUVIMED