Study on Brain Functional Networks and Wearable Device Monitoring in Children with SeLECTs and High Spike-Wave Index (SWI >50%)

Minghao Xu¹Xiaoxuan Li¹Yifan Fu¹Dinghan Hu²Yang Zhang³Wei Wang¹Yaotian Gao¹Keyi Lin¹Bin Yang⁴Tao Jiang^1*Jiuwen Cao^5*

• ¹Anhui Public Health Clinical Center, Anhui Medical University, Hefei, China
• ²Machine Learning and I-health International Cooperation Base of Zhejiang Province, and the Artificial Intelligence Institute, Hangzhou Danzi University, 310000 Zhejiang, China, Hangzhou, China
• ³Department of Neurosurgery, The First Affiliated Hospital of Anhui University of Traditional Chinese Medicine, 230000 Hefei, Anhui, China, Hefei, China
• ⁴Department of Neurology, Anhui Province Children’s Hospital, 230000 Hefei, Anhui, China, Hefei, China
• ⁵Machine Learning and I-health International Cooperation Base of Zhejiang Province, and the Artificial Intelligence Institute, Hangzhou Dianzi University, Zhejiang, China,, Hangzhou, China

Background: Self-limited epilepsy with centrotemporal spikes (SeLECTs) represents a common idiopathic focal epilepsy syndrome in childhood. Although most patients demonstrate a favorable prognosis, some patients develop ESES. ESES is associated with poorer neuropsychological prognosis. This association challenges the "benign" classification of SeLECTs. Currently, the diagnostic threshold for ESES remains controversial. Moreover, traditional long-term video-EEG monitoring presents certain limitations. Methods: The research utilizes the "Biovital-P1" software integrated with Oppo smart bands to collect multimodal physiological signals. Simultaneously, a 21-channel digital EEG system acquires electroencephalographic data. The study constructs brain networks through DTF analysis. Additionally, it performs preprocessing and feature extraction on multimodal physiological signals (ACC, EDA, PPG). Results: The results demonstrate strong functional connectivity in the centrotemporal region in all frequency bands and the delta band. However, as SWI levels increase, the brain network's global and local efficiency significantly reduces. Analyzing multimodal physiological signals reveals statistically significant differences in ACC and PPG signal time-domain features (Maximum, Minimum, Peak) among different SWI groups. Multiple characteristic parameters of EDA signals Sample et al. Running Title also show significant intergroup differences. Notably, EDA signals exhibit excellent sensitivity in reflecting stress responses of the autonomic nervous system. The characteristic features of EDA signals demonstrate a significant negative correlation with SWI levels.