CMEpiNet: Complex-Valued Multimodal Epilepsy Detection Network Model
Sensors · 02.07.2026 · 26 (13) · 4186
DOI: 10.3390/s26134186 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, online erschienen am 02.07.2026. Verfasst von Tianyi Su und 3 weiteren Autor:innen. Beteiligte Einrichtung: Department of Electrical and Information Engineering, Shandong University of Science and Technology, Jinan 250031, China. Hauptthemen laut MEDLINE: Epilepsy. Die Arbeit nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Innovation Capacity Improvement Project for Technology-based SMEs in Shandong Province, Youth Innovation Team Development Plan for Universities in Shandong Province, College Students’ Innovative Entrepreneurial Training Plan Program Project.
Abstract (Originalsprache)
Existing seizure detection methods cannot fully exploit the spatiotemporal features of multimodal signals. They also fail to capture deep associations among cross-modal features. This limits their ability to learn unified representations of spatiotemporal dependencies. This work proposes CMEpiNet (Complex-valued Multimodal Epilepsy detection Network model) to address this issue. CMEpiNet first uses complex-valued convolutions for feature extraction. It explicitly models phase synchronization, phase shifts, and cross-frequency coupling. Thus, EEG, ECG, and EMG features are represented in the complex-valued domain. During feature fusion, CMEpiNet uses a two-level semantic alignment-based fusion method. It applies cross-modal consistency constraints in a shared alignment space. It also performs distribution-level alignment in an epilepsy-related semantic latent space. These operations ensure the consistency of multimodal features in the global semantic structure. Finally, CMEpiNet uses a spatial attention-guided 3D convolutional classifier. The classifier jointly models the temporal, feature, and modality dimensions. Experimental results on the SeizeIT2 dataset show that CMEpiNet improves seizure detection sensitivity, reduces the false alarm rate, and maintains stable performance under perturbations.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Electrical and Information Engineering, Shandong University of Science and Technology, Jinan 250031, China.
- School of Physical Education and Health Linyi, Linyi University, Linyi 276012, China.
- School of Information Science and Engineering, Linyi University, Linyi 276012, China.
ORCID-Profile: Haifeng Wang
Förderung
- Innovation Capacity Improvement Project for Technology-based SMEs in Shandong Province
- Youth Innovation Team Development Plan for Universities in Shandong Province
- College Students’ Innovative Entrepreneurial Training Plan Program Project
Zitieren
APA
Su, T., Zhu, H., Chen, S., & Wang, H. (2026). CMEpiNet: Complex-Valued Multimodal Epilepsy Detection Network Model. Sensors, 26(13), 4186. https://doi.org/10.3390/s26134186
Vancouver
Su T, Zhu H, Chen S, Wang H. CMEpiNet: Complex-Valued Multimodal Epilepsy Detection Network Model. Sensors. 2026;26(13):4186. doi:10.3390/s26134186
BibTeX
@article{Su2026,
author = {Su, Tianyi and Zhu, Haiyan and Chen, Shuai and Wang, Haifeng},
title = {CMEpiNet: Complex-Valued Multimodal Epilepsy Detection Network Model},
journal = {Sensors},
year = {2026},
volume = {26},
number = {13},
pages = {4186},
doi = {10.3390/s26134186}
}
EUVIMED