On Functional Connectivity And Effectivity Connectivity Network In Minimal Hepatic Encephalopathy Using Complex Network Methods

With the concepts of small-world network and scale-free netwok proposed in complex networks,complex networks has been developed and used in many research fields, such as sociology, computerscience, and biology. Also, complex networks can provide new perspective for the research fieldsabove, including the research of neuroscience. Complex networks are able to quantify the analyses ofbrain network and brain functions and brain damages in strict mathematical views. Previous studies invarious brain diseases and brain functions showed that complex networks were useful for braindisease and brain function studies.Patients with liver disease could have high event rate of minimal hepatic encephalopathy whichmight develop into clinical hepatic encephalopathy if no actions were token. Hence, it is of greatimportance to study minimal hepatic encephalopathy. Many Studies had shown that brain functionalconnectivity network in patients with minimal hepatic encephalopathy were abnormal. However, noresearch has been performed to study brain networks via complex network methods. The main worksof this thesis were as follows. Firstly, we built the brain functional network of patients with minimalhepatic encephalopathy based on the BOLD-fMRI data and analyzed the brain functional network bystatistical parametric of complex networks, such as clustering coefficient, average path length,small-world characteristics. Results showed that the small-world characteristics and the normalizedclustering coefficient normalized changed in patients with minimal hepatic encephalopathy, whilenormalized average path length was nearly unchanged. These changes had correlations withneuropsychological tests of patients with minimal hepatic encephalopathy, which showed that thesefeatures could be used to diagnose minimal hepatic encephalopathy. Secondly, we partitioned thebrain functional network of patients with minimal hepatic encephalopathy by the communitypartitioning algorithm in complex networks. Results showed that patients with minimal hepaticencephalopathy had lower maximum modularity, and their networks were divided into morecommunities, which were related with the impediments of cooperative communication between thebrain regions in patients with minimal hepatic encephalopathy. We also found that the basalganglia-thalamus-cortical loop was impaired in patients with minimal hepatic encephalopathy, whichcould be pathologically associated with minimal hepatic encephalopathy. Finally, we developed braineffectivity network measurements in patients with minimal hepatic encephalopathy based on theGranger Causality and analyzed the effectivity network using characteristics, eg. average path length, indegree and outdegree. Results showed that the information outflow of the basal ganglia and thalamregions in left hemisphere were impaired, while the average path length of the patients’ networks wasnot, indicating that there could be a more power argument in favor of the impediment of basal gangliaand thalam regions in patients with minimal hepatic encephalopathy.In this thesis, we studied the brain network of patients with minimal hepatic encephalopathyusing complex networks, and results showed that complex networks could be used to study minimalhepatic encephalopathy and could play an improtant role in future diagnosis or the pathologicalrevelation of minimal hepatic encephalopathy.