Study On Abnomalities In Topology And Correlation Of Major Depressive Disorder Patients’ Brain Network

Depressive disorder is a mental disease whose clinical performances are anxiety, depression, pessimism and despair. Especially, major depressive disorder(MDD) is the most serious and people with major depressive disorder have the tendency to suicide. Human brain system is one of the most complex systems, complexity is not only reflected in the hundreds of millions of neurons are connected, but also int the synergies on thought, emotion, cognition and other advanced mode. Recently, researches used FMRI get brain images, construct of a brain network, and applied complex network methods to study brain networks, expect to find potential network topology relationship. Complex network theory allows us from a different perspective to the study of human brain function mechanism. It provides a new theoretical basis for the treatment and prevention of depression.This paper does the interdisciplinary research based on theories of complex network and theories of brain science, apply it on solving actual problem that is treatment and prevention on MDD. Explore the methods of building brain functional network. We use the network indicators were compared to explore the differences between groups, including global efficiency, local efficiency, nodal efficiency, edge betweenness, and modular properties. It can provide a reliable theoretical basis for the treatment of depression.In conclusion, five things we have done are as followed:(1)Construct functional brain network.This paper used FMRI to get the brain images and do image preprocessing. The brain is segmented into 90 brain regions by AAL template. The adjacency matrix of the functional brain network is gained by Pearson correlation coefficient between regions which has a corresponding value of average time sequence.(2)Analyze the network on its features and efficiency.Through computing the global efficiency and local efficiency in MDD network, normal people’s brain network, random network and regular network, we uncover the functional network has the properties of small world network and economic nature.(3) Analyze the nodes and edges efficiency. In order to investigate the difference between patients’ brain network and normal control’s network, we observe the node and edge betweenness in both networks and use two sample T test. Results show that patients with MDD has a rise on node efficacy in the marginal regions, reduction on node efficacy in the executive regions and a great reduction on edge efficiency between marginal regions and executive regions. In detail, the executive brain regions is in charge of regulation and marginal brain regions is responsible. This means that the patient with MDD has low ability on adjust negative feelings and increased the process on negative thoughts. These two actions would lead the patients has excess negative emotions and to be depressive. Thus, enhancing the activity in executive functional regions and inhibiting the activity in marginal regions could help a lot on treatment.(4) Analyze the community structure, k-core and closeness in network. Besides on analysis on nodes and edges, the community structure, k-core and closeness are also studied. We apply Newman module algorithm and CNN module partition algorithm to observe the community structure and compute the k-core and closeness. The final results tell that functional brain network has remarkable modularity than random network. The patients’ brain network has a drop on k-core and closeness between executive functional regions and language functional regions. It in further explains patients’ language function decline and patients have difficulty in communicating and like to be self-closed. All descriptions are same with external manifestations in clinical observation.(5)Analyze the correlations among global efficiency, age, gender, education and medical time. Pearson correlation coefficient is implemented to study the relation among global efficiency, age, gender, education and medical time. We figure out that global efficiency of the brain functional network is negatively correlated with age and medical time, positively correlated with gender and education. This can help us give suggestions on these kinds of people how to prevent getting MDD.