A Study On The Topological Properties And Functional Connection Patterns Of Brain Networks In Patients With Alzheimer's Disease

Alzheimer's disease(AD)is a neurodegenerative disease characterized by cognitive decline.With the development of neuroimaging,network analysis based on graph theory has been used to characterize the internal structure of human brain network.Although a large number of studies have found abnormal brain network segregation and integration in patients with AD,the modular organization of the brain network in patients with AD is still largely unknown.The functional connectivity network(FCN)and high-order functional connectivity network(HOFC)in patients with AD are usually detected separately.The results are different.Therefore,based on the traditional functional brain network,this paper constructs the higher-order brain network,examines its modular structure,and proposes a new detection index,which provides a new direction for understanding the pathophysiological mechanism of AD.The specific work is as follows:Firstly,a high-order brain network difference detection algorithm based on spectral optimization is proposed.The time series of subjects' brain regions were obtained and Pearson's correlation was calculated to obtain the low-order functional brain network.And the correlation of topology features of low-order networks is calculated to obtain high-order brain networks,and the network modules are identified by spectral optimization algorithm.In addition,the abnormal brain network was analyzed based on graph theory to find out the characteristics of abnormal brain network.The results showed that,compared with healthy controls(HC)group,the segregation of FCN and HOFCN(increased clustering coefficient value and reduced connection number in default mode network)and integration(increased characteristic path length and normalized characteristic path length)were destroyed.As for the detection of brain network modular organization,this paper found that compared with HC group,the modular integrity of FCN and HOFCN in AD patients was damaged,which was mainly reflected in the significant reduction of characteristic path length in the default mode network.In the lateral frontal lobe region of HOFCN,this paper also found a significant reduction in the number of intra-modular connections.In addition,HOFCN is not completely consistent with traditional FCN in connectivity,modular number,and modular topology,which also reveals the additional information provided by high-order networks.Secondly,a difference detection algorithm based on the coupling of low-order and high-order brain networks is proposed.The low-order and high-order brain networks were constructed in this paper.And the brain network was divided into modules by using the spectraloptimization algorithm,and the network topology attributes were calculated at the global,modular and nodal levels to evaluate the coupling of FCN and HOFCN.In addition,based on the analysis of different levels,abnormal brain network characteristics were found.The results showed that,compared with HC group,the coupling between functional connectivity(FC)and high-order functional connectivity(HOFC)in DMN was significantly increased,and the inter-network coupling of characteristic path length and normalized characteristic path length in the module significantly decreased.Finally,at the nodal level,this paper found that compared with HC group,the nodal coupling degree of AD patients in bilateral dorsolateral superior frontal gyrus,middle frontal gyrus,triangular inferior frontal gyrus and medial superior frontal gyrus was significantly reduced.