Small-world Network for Investigating Functional Connectivity in Bipolar Disorder: A Functional Magnetic Images (fMRI) Study

Abstract

Small-world network is a highly clustered system but with small mean path length between networks which allow the information transferred with high efficiency. The human brain can be considered as a sparse, complex network modeled by the small-world properties. Once the brain network was disrupted by disease, the small-world properties would be altered to manifest that the information integration was inefficiency and the network was loosely organized. The aim of this study is to investigate the difference of small-world properties of brain functional network derived from resting-state functional magnetic resonance imaging (fMRI) between the healthy subjects and the patients with Bipolar disorder (BD). The functional MRI data was acquired from 5 healthy subjects and 5 patients with Bipolar disorder. All images of each subject were parcellated into 90 cortical and sub-cortical regions which were defined as the nodes of the network. The functional relations between the 90 regions were estimated by the frequency-based mutual information followed by thresholding to construct a set of undirected graphs. Small-world properties, such as the degree and strength of the connectivity, clustering coefficient of connections, mean path length among brain regions, global efficiency and local efficiency, are examined between any pair of functional areas. Our findings indicated that, in comparison with the control subjects, the BD patients presented smaller values of the degree, the strength of the connectivity and the clustering coefficient of connections, whereas larger values of mean path length among brain regions. This suggested the reduced global and local efficiency of the small-world properties for BD patients. In addition, the small-world properties of BD patients were altered significantly in some regions in the frontal lobes and limbic system which were in good agreement with the dysfunction connectivity reported by the previous literatures in the study of bipolar disorder.