WaveCrossNet: interpretable cross-modal automatic sleep staging for OSA patients
Biomedical Physics & Engineering Express · 01.08.2026 · 12 (4) · 045065
DOI: 10.1088/2057-1976/ae93bb ↗Kurz gefasst
Fachartikel in der Zeitschrift Biomedical Physics & Engineering Express, online erschienen am 14.08.2026. Verfasst von Ruomeng Quan und 8 weiteren Autor:innen. Beteiligte Einrichtung: School of Physics and Information Technology, Shaanxi Normal University, Xi'an, Shaanxi 710119, People's Republic of China. Hauptthemen laut MEDLINE: Sleep Apnea, Obstructive, Sleep Stages, Signal Processing, Computer-Assisted. Die Arbeit nennt 45 Literaturangaben. Gefördert von: Xianyang Science and Technology Bureau Project, Natural Science Basic Research Plan in Shaanxi Province of China.
Abstract (Originalsprache)
Obstructive sleep apnea (OSA) patients show highly fragmented sleep with frequent stage transitions, challenging existing automatic sleep staging in feature extraction, multimodal correlation modeling, and interpretability. Current methods rely on fixed-threshold denoising and standard positional encoding, suffering from obvious module stacking and a lack of design tailored to OSA’s non-stationary signals. To address these issues, this paper proposes WaveCrossNet, an enhanced Transformer-based model for OSA sleep staging. Instead of simple module stacking, WaveCrossNet embeds adaptive wavelet thresholding and dual-channel convolutional neural network into Transformer’s end-to-end temporal learning pipeline. First, an energy-adaptive subband thresholding mechanism selectively denoises while preserving transient events such as micro-arousals. Second, position-aware attention with dynamic cross-modal fusion explicitly models complementary patterns of electroencephalography and electrooculography across N 1, REM and other sleep phases. Third, a hard example mining loss alleviates class imbalance, where minority stages like N 1 account for less than 5% of OSA data. Attention weight visualization links feature correlations to physiological events (e.g. slow eye movements in N 1, high-amplitude slow waves in N 3), enhancing interpretability. Evaluated on the Sleep-EDFx healthy dataset and 200 Sleep Heart Health Study OSA records with subject-independent splits, WaveCrossNet achieves competitive metrics on healthy subjects and obtains state-of-the-art performance on OSA patients with an overall accuracy of 84.9%. By jointly optimizing denoising, representation learning, and temporal modeling, WaveCrossNet provides an interpretable, end-to-end solution for accurate OSA sleep staging.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- School of Physics and Information Technology, Shaanxi Normal University, Xi'an, Shaanxi 710119, People's Republic of China.
ORCID-Profile: Ruomeng Quan, Jiaxin Tai, Mengyuan Liu, Yanyan Xiao, Jingjing Yu, Jin Li, Ning Qiang
Förderung
- Xianyang Science and Technology Bureau Project
- Natural Science Basic Research Plan in Shaanxi Province of China
Zitieren
APA
Quan, R., Tai, J., Liu, M., Zhang, X., Xiao, Y., Yu, J., Hu, J., Li, J., & Qiang, N. (2026). WaveCrossNet: interpretable cross-modal automatic sleep staging for OSA patients. Biomedical Physics & Engineering Express, 12(4), 045065. https://doi.org/10.1088/2057-1976/ae93bb
Vancouver
Quan R, Tai J, Liu M, Zhang X, Xiao Y, Yu J, et al. WaveCrossNet: interpretable cross-modal automatic sleep staging for OSA patients. Biomedical Physics & Engineering Express. 2026;12(4):045065. doi:10.1088/2057-1976/ae93bb
BibTeX
@article{Quan2026,
author = {Quan, Ruomeng and Tai, Jiaxin and Liu, Mengyuan and Zhang, Xu and Xiao, Yanyan and Yu, Jingjing and Hu, Jing and Li, Jin and Qiang, Ning},
title = {WaveCrossNet: interpretable cross-modal automatic sleep staging for OSA patients},
journal = {Biomedical Physics & Engineering Express},
year = {2026},
volume = {12},
number = {4},
pages = {045065},
doi = {10.1088/2057-1976/ae93bb}
}
EUVIMED