Lu Meili, Guo Zhaohua, Gao Zicheng
School of Information Technology Engineering, Tianjin University of Technology and Education, Tianjin, China.
Front Hum Neurosci. 2023 Dec 20;17:1295326. doi: 10.3389/fnhum.2023.1295326. eCollection 2023.
The objective of this study was to explore the distributed network effects of intracranial electrical stimulation in patients with medically refractory epilepsy using dynamic functional connectivity (dFC) and graph indicators.
The time-varying connectivity patterns of dFC (state-based metrics) as well as topological properties of static functional connectivity (sFC) and dFC (graph indicators) were assessed before and after the intracranial electrical stimulation. The sliding window method and -means clustering were used for the analysis of dFC states, which were characterized by connectivity strength, occupancy rate, dwell time, and transition. Graph indicators for sFC and dFC were obtained using group statistical tests.
DFCs were clustered into two connectivity configurations: a strongly connected state (state 1) and a sparsely connected state (state 2). After electrical stimulation, the dwell time and occupancy rate of state 1 decreased, while that of state 2 increased. Connectivity strengths of both state 1 and state 2 decreased. For graph indicators, the clustering coefficient, k-core, global efficiency, and local efficiency of patients showed a significant decrease, but the brain networks of patients exhibited higher modularity after electrical stimulation. Especially, for state 1, there was a significant decrease in functional connectivity strength after stimulation within and between the frontal lobe and temporary lobe, both of which are associated with the seizure onset.
Our findings demonstrated that intracranial electrical stimulation significantly changed the time-varying connectivity patterns and graph indicators of the brain in patients with medically refractory epilepsy. Specifically, the electrical stimulation decreased functional connectivity strength in both local-level and global-level networks. This might provide a mechanism of understanding for the distributed network effects of intracranial electrical stimulation and extend the knowledge of the pathophysiological network of medically refractory epilepsy.
本研究的目的是使用动态功能连接（dFC）和图指标来探索颅内电刺激对药物难治性癫痫患者的分布式网络效应。
在颅内电刺激前后评估dFC的时变连接模式（基于状态的指标）以及静态功能连接（sFC）和dFC的拓扑特性（图指标）。使用滑动窗口法和K均值聚类分析dFC状态，其特征在于连接强度、占有率、停留时间和转换。通过组统计检验获得sFC和dFC的图指标。
DFC被聚类为两种连接配置：强连接状态（状态1）和稀疏连接状态（状态2）。电刺激后,状态1的停留时间和占有率降低，而状态2的停留时间和占有率增加。状态1和状态2的连接强度均降低。对于图指标，患者的聚类系数、k核、全局效率和局部效率均显著降低，但电刺激后患者的脑网络表现出更高的模块化。特别是，对于状态1，额叶和颞叶内部以及两者之间刺激后的功能连接强度显著降低，这两者均与癫痫发作起始有关。
我们的研究结果表明，颅内电刺激显著改变了药物难治性癫痫患者大脑的时变连接模式和图指标。具体而言，电刺激降低了局部水平和全局水平网络中的功能连接强度。这可能为理解颅内电刺激的分布式网络效应提供一种机制，并扩展对药物难治性癫痫病理生理网络的认识。
