Construction And Analysis Of Abnormal Brain Functional Networks

Brain functional network research is of great significance to understand the brain information processing and knowledge expression of neurophysiological mechanism.The modern brain imaging techniques and complex network theory provides the necessary basis for the study on brain networks.In this work,Alzheimer's disease(AD)brain electrical signals of resting state have been applied to construct abnormal brain networks.Analysis of the topology change will facilitates to understand the pathogenesis and provide auxiliary for diagnosis.The complex network theory is applied to EEG series to investigate the nonlinear characteristic and the cognitive function of the brain with Alzheimer's disease(AD).The Limited penetrable visibility graph(LPVG)and phase space method are used to map single EEG series into networks,and investigate the underlying chaotic system dynamics of the AD brain.By further calculation of Pearson correlation coefficient of Clustering coefficient sequences,corresponding brain functional networks are constructed using the theory of graph theory to extract the network topology parameters.With the analysis of each frequency bands of AD abnormal brain functional networks,significant differences are found mainly in alpha band and brain region anomalies mainly appear at the top,right occipital and temporal region.Synchronization likelihood(SL)and fuzzy synchronization likelihood(FSL)are applied to describe the nonlinear synchronization and construct the brain functional networks.This article first applied simulation data generated by the Henon model to verify the effectiveness of the proposed algorithm and compare their advantages and disadvantages.The AD brain functional networks show loss of connection,which results in the decrease of Graph Index Complexity and small world properties.Through a variety of methods of brain network building and analysis,this study obtains Alzheimer's effects on brain functional networks topology and found that the adopted networks can detect network characteristic parameter changes on Alzheimer's disease diagnosis and evaluation of rehabilitation.