Altered Dynamic Functional Network Connectivity in Frontal Lobe Epilepsy.

ABSTRACT

Frontal lobe epilepsy has recently been associated with disrupted brain functional connectivity; variations among various resting-state networks (RSNs) across time remains largely unclear. This study applied dynamic functional network connectivity (dFNC) analysis to investigate functional patterns in the temporal and spatial domains of various functional systems in FLE. Resting-state fMRI data were acquired from 19 FLE patients and 18 controls. Independent component analysis was used to decompose RSNs, which were grouped into seven functional systems. Sliding windows and clustering approach were used to identify the dFNC patterns. Then, state-specific connectivity pattern and dynamic functional state interactions (dFSIs) were evaluated. Compared with healthy controls, FLE patients exhibited decreased dFNC in almost all four patterns, changes that were mostly related to the frontoparietal system, suggesting a disturbed communication of the frontoparietal system with other systems in FLE. Additionally, regarding the fundamental connectivity pattern (state 3 in this study), FLE showed decreased time spent in this state. Moreover, the duration positively correlated with seizure onset. Furthermore, significantly reduced dynamic connections in this state were observed in the frontoparietal system linked to the cerebellar and subcortical systems. These findings imply abnormal fundamental dynamic interactions and dysconnectivity associated with the subcortical and cerebellar regulation of dysfunctions in frontoparietal regions in FLE. Finally, based on the developed FSI analysis, temporal dynamic abnormalities among states were observed in FLE. Therefore, this altered dynamic FNC extended our understanding of the abnormalities in the frontoparietal system in FLE. The dynamic FNC provided novel insight into the fundamental pathophysiological mechanisms in FLE.