Analysis of brain network differences in the active, motor imagery, and passive stoke rehabilitation paradigms based on the task-state EEG

•A range of graph theory indices were utilized to analyze the brain networks of both stroke patients and healthy subjects.•The differences in brain network connectivity across the active, motor imagery, and passive task paradigms were compared.•Both ipsilateral and contralateral motor areas were reorganized in stroke patients during the MI task.•Compared to active and passive tasks, motor imagery demonstrates a greater capacity to promote brain reorganization.•This study proposed a potential explanatory mechanism for the role of motor imagery in promoting stroke rehabilitation. Different movement paradigms have varying effects on stroke rehabilitation, and their mechanisms of action on the brain are not fully understood. This study aims to investigate disparities in brain network and functional connectivity of three movement paradigms (active, motor imagery, passive) on stroke recovery. EEG signals were recorded from 11 S patients (SP) and 13 healthy controls (HC) during fist clenching and opening tasks under the three paradigms. Brain networks were constructed to analyze alterations in brain network connectivity, node strength (NS), clustering coefficients (CC), characteristic path length (CPL), and small-world index(S). Our findings revealed increased activity in the contralateral motor area in SP and higher activity in the ipsilateral motor area in HC. In the beta band, SP exhibited significantly higher CC in motor imagery (MI) than in active and passive tasks. Furthermore, the small world index of SP during MI tasks in the beta band was significantly smaller than in the active and passive tasks. NS in the gamma band for SP during the MI paradigm was significantly higher than in the active and passive paradigms. These findings suggest reorganization within both ipsilateral and contralateral motor areas of stroke patients during MI tasks, providing evidence for neural restructuring. Collectively, these findings contribute to a deeper understanding of task-state brain network changes and the rehabilitative mechanism of MI on motor function.