Abnormal topology of brain functional networks in unipolar depression and bipolar disorder using optimal graph thresholding

Zhinan Yu,Jiaolong Qin,Xinyuan Xiong,Fengguo Xu,Jun Wang,Fengzhen Hou,Albert Yang
DOI: https://doi.org/10.1016/j.pnpbp.2019.109758
2020-01-01
Abstract:Two popular debilitating illness, unipolar depression (UD) and bipolar disorder (BD), have the similar symptoms and tight association on the psychopathological level, leading to a clinical challenge to distinguish them. In order to figure out the underlying common and different mechanism of both mood disorders, resting-state functional magnetic resonance imaging (rs-fMRI) data derived from 36 UD patients, 42 BD patients (specially type I, BD-I) and 45 healthy controls (HC) were analyzed retrospectively in this study. Functional brain networks were firstly constructed on both group and individual levels with a density 0.2, which was determined by a network thresholding approach based on modular similarity. Then we investigated the alterations of modular structure and other topological properties of the functional brain network, including global network characteristics and nodal network measures. The results demonstrated that the functional brain networks of UD and BD-I groups preserved the modularity and small-worldness property. However, compared with HC, reduced number of modules was observed in both patients' groups with shared alterations occurring in hippocampus, para hippocampal gyrus, amygdala and superior parietal gyrus and distinct changes of modular composition mainly in the caudate regions of basal ganglia. Additionally, for the network characteristics, compared to HC, significantly decreased global efficiency and small-worldness were observed in BD-I. For the nodal metrics, significant decrease of local efficiency was found in several regions in both UD and BD-I, while a UD-specified increase of participant coefficient was found in the right paracentral lobule and the right thalamus. These findings may contribute to throw light on the neuropathological mechanisms underlying the two disorders and further help to explore objective biomarkers for the correct diagnosis of UD and BD.
What problem does this paper attempt to address?