Brain network analysis in Parkinson’s disease patients based on graph theory

Abstract

AbstractDevelopment of Parkinson’s disease causes functional impairment in the brain network of Parkinson’s patients. The aim of this study is to analyze brain networks of people with Parkinson’s disease based on higher resolution parcellations and newer graphical features. The topological features of brain networks were investigated in Parkinson’s patients (19 individuals) compared to healthy individuals (17 individuals) using graph theory. In addition, four different methods were used in graph formation to detect linear and nonlinear relationships between functional magnetic resonance imaging (fMRI) signals. The functional connectivity between the left precuneus and the left amygdala, as well as between the vermis 1-2 and the left temporal lobe was evaluated for the healthy and the patient groups. The difference between the healthy and patient groups was evaluated by non-parametric t-test and U-test. Based on the results, Parkinson’s patients showed a significant decrease in centrality criterion compared to healthy subjects. Furtheremore, changes in regional features of brain network were observed. There was also a significant difference between the two groups of healthy subjects and Parkinson’s patients in different areas by applying centrality criterion and the correlation coefficients. The results obtained for topological features indicate changes in the functional brain network of Parkinson’s patients. Finally, similar areas obtained by all three methods of graph formation in the evaluation of connectivity between paired regions in the brain network of Parkinson’s patients increased the reliability of the results.