Dynamic Functional Connectivity in Lateralization of Temporal Lobe Epilepsy

Background: Epilepsy is conceptualized as a network disease that characterized by recurrent and temporary brain dysfunctions due to discharges of interconnected groups of neurons. The network can be characterized by inter-regional functional connectivity, i.e., the correlations of blood oxygen level-dependent (BOLD) signals between any two functional regions in the brain. However, since the BOLD signal is inherently non-stationary, the functional connectivity is evidenced to be varying over time. Thus, the correlations computed at different time points across the measurements could preserve significant variabilities.By considering dynamic characteristics of the functional network and using graph theoretical analyses, we aimed to obtain transition of network properties that account for the lateralityin temporal lobe epilepsy (TLE) patients. Methods: Twenty patients with left-TLE and 15 patients with right-TLE underwent resting-state functional MRI. We calculated the dynamic functional connectivity (the time-varying inter-regional correlation values) using a sliding window technique. We also calculated static functional connectivity for the purpose of comparison with dynamicresults. We measured six graph theoretical characteristics including clustering coefficient, degree centrality, betweenness centrality, node neighbor’s degree, closeness centrality, and page rank. Findings: A multilevel mixed effect linear regression model suggesteda significant increase in clustering coefficient, betweenness centrality and node neighbor’s degree and a significant decrease degree centrality and closeness centrality in patients with left TLE comparedto thepatients with right TLE (P < 0.05). Page rank feature, however, did not show any significant difference between two groups. The results demonstrated that the static connectivity analysis failed to separate the left and right TLE patients by clustering coefficient, node neighbor’s degree, and page rank. Conclusion: We conclude that accounting for the non-stationarity characteristics of functional connectivity accompanied with the graph theoretical measures can be a prerequisite in the search for potential connectivity-derived lateralization biomarkers in TLE.

کلیدواژه ها:

Dynamic Functional Connectivity ، TLE ، Graph Theory ، Lateralization ، fMRI