Study Of Neural Mechanisms In Loneliness Based On Diffusion Tensor Imaging

Loneliness is a kind of subjective psychological feeling or experience with lack ofsatisfying relationships that they desired. If loneliness lasts for a long time, it wouldaffect physical and mental health, even lead to cognitive dysfunction. Though a largenumber of fMRI studies have demonstrated that functional activation was decreased inloneliness related regions. Individual structural differences still remain unknown. Thesubtle changes in brain white matter (WM) fiber microstructure can be detected in vivoby diffusion tensor imaging (DTI) technology via water molecules move dispersively.DTI provides a new technical support for the study of brain structure regions inloneliness. Here, we utilize DTI technology to investigate the neural mechanisms ofloneliness based on WM in the region of interest (ROI) and the topological organizationof human anatomical networks. The main works of this thesis include:1. Exploring firstly the relationship between fractional anisotropy (FA) of WM inthe ventral attention network (VAN) and loneliness. Mean FA values between each pairof regions (inferior fronto-occipital fasciculus, external capsule, inferior longitudinalfasciculi) are calculated. Pearson correlation is used to compute the relationshipbetween VAN related regions and loneliness scale. We found the higher of lonelinessscore is rated, the lower of the anisotropy of their tissues obtained. The significantnegative correlation of mean FA values in WM of VAN illustrates that the weakenstructure connections in VAN related regions may be related to the decrease ofcognitive function.2. Investigating WM structure networks in loneliness. DTI tractography is used totrack the whole WM fibers. We first reconstruct the human brain WM network inloneliness. Graph-theoretical approaches are applied to investigate topologicalorganization and nodal properties of the network in loneliness. The results reveal thatthe WM networks of all participants show a small-world topology. However, thesmall-world properity, normalized clustering coefficient, and global efficiency aresignificantly decreased. The normalized characteristic path length is significantlyincreased in the high loneliness compared to low loneliness. Moreover, the localefficiency in attention network, sensorimotor networks, and emotional processingnetwork are significantly negatively correlated with loneliness score.Furthermore, we investigate the white matter fiber properties in emotion processingrelated areas. The result shows a significant negative correlation between the FA values of emotional processing related areas (fusiform gyrus, ventral striatum) and lonelinessscore. Our findings may provide important information on the neurocognitiveprocessing in loneliness.