Research On Modeling And Dynamic Connectivity Of Multilayer Brain Network

In recent years,research on brain functional connectivity of complex networks has become a hot topic,and using f MRI data to construct brain networks has become an important method for human brain research.The traditional research paradigm was to abstract the brain into a single-layer network and focused on the correlation of the BOLD effect of each brain region over the entire time period.However,the singlelayer network has unavoidable defects that the individual point in the entire time period cannot be considered as a whole.Regardless of the resting state or the task state,the brain will show different functional states and neural activity patterns at different moments.The multilayer temporal network model can take the time factor into account to study the evolution of brain networks over time.Based on this,this study hopes to explore the dynamic functional connectivity of the brains of HC and JME patients by constructing a multilayer temporal network model with computer technology.The main research contents are as follows:Single-layer network models are simple,tractable and well-established for studying functional connectivity of the resting state brain.Multilayer network can represent richer information and better simulate the dynamics of the brain.In order to compare and study the characteristics of the two network models,we selected HC as the research object.A single-layer network model and a multilayer network model were constructed respectively.Community detection was carried out.The differences of their module division,the connections between network metrics and the classification results of whole brain nodes were analyzed.The results showed that the quantity of modules and the quality of modules divided by the multilayer network are significantly higher than those of the single-layer network.Moreover,the multilayer network has unique network metrics in quantifying the functional separation,functional integration and importance of brain regions.By calculating these metrics,multilayer network can identify more important brain regions that are different from the single-layer network.These results suggest that multilayer networks are more sensitive to capturing dynamic changes in the resting state brain.Based on the multilayer network model,the dynamic changes in the brain functional network can be described more comprehensively and accurately.This study further uses this model to analyze the brain network of JME patients in order to better understand the changes in the brain functional network of JME patients.This study selected JME patients and corresponding HC as research objects.A multilayer network model was constructed and community detection was performed.The recruitment,integration,flexibility,promiscuity and core-peripheral system were calculated.The results showed that the module quality of the JME patient group was significantly decreased.The recruitment was decreased and the integration,flexibility and promiscuity were increased in some functional networks and brain regions of the JME patient group.The left insula and left cuneus were core regions specific to the JME group,while most of the peripheral systems specific to the JME group were distributed in the prefrontal cortex and hippocampus.The flexibility of the opercular part of the inferior frontal gyrus was significantly correlated with the severity of JME symptoms.These results indicated that the dynamic community structure of the JME patients is abnormal.