The Research Of Brain Network Topology Changes In Patients With Post-stroke Depression

Objective: To explore the possible changes of the network topological attributes of the whole white matter structure in patients with post-stroke depression(PSD)by combining of DTI technology and the method of graph theory to find the PSD brain network image markers and to provide objective diagnostic indicators for early diagnosis and curative effect evaluation.Methods: Forty-eight patients with acute ischemic stroke who underwent ischemic stroke within 2 weeks were divided into PSD group and Non-PSD group according to the DSM-IV standard.Among them,HAMD scores were requested> 8 in PSD group and <8 in Non-PSD,subjects were scanned by 3.0T SIEMENS MRI conventional and diffusion tensor imaging(DTI).First,the PANDA software was used to pre-treat DTI data from all patients,Then,the entire brain was divided into 90 regions by using Automated anatomical labeling(AAL).PANDA was used to track the whole brain white matter.Finally,gretna software was used to analyze the white matter network attributes of the two groups.And the two samples of t test.The main concerns of this study are as follows:(1)Based on the global study of brain network,the small world attributes(sigma),clustering coefficient(Cp),standard feature path length(Lambda)of PSD group and NonPSD group,),And the characteristic path length(Lp)for two-sample t-test.(2)Based on the study of network efficiency,including the PSD group and NonPSD group global efficiency(Eglob),local efficiency(Eloc)for two-sample t test.(3)Based on the study of network local efficiency,including the PSD group and Non PSD group,each patient's 90 brain regions were clustered coefficient and local efficiency values were two-sample t-test.Results:(1)PSD group and Non-PSD group had small world attributes,but the small-world attribute of PSD group was smaller than that of Non-PSD group(P <0.05),and the PSD cluster coefficient(Cp)(P <0.05).In addition,the normalized clustering coefficient(Gamma)of the PSD group,the standard characteristic path length(Lambda),and the characteristic path length Lp were lower than those of the Non-PSD group(P <0.05),and the difference was statistically significant(P <0.05).(2)There were significant differences in the global efficiency(Eglob)and local efficiency(Eloc)of the network efficiency of the PSD group compared with the non-PSD group(P <0.05))(3)Efficacy changes in the brain area: left orbital frontal gyrus,right orbital frontal gyrus,left medial frontal gyrus,left anterior cingulate gyrus and paravertebral brain,right anterior cingulate and paracellus,Back hippocampus,right hippocampus,right thalamus,right temporal gyrus nodes increased efficiency.Node clustering coefficient(Cp)to change the brain area: the right side of the olfactory cortex node efficiency increased,the left dorsal lateral volume back,right dorsal lateral frontal gyrus,right frontal frontal gyrus,the middle of the left back,left the island cover The right hippocampus,the right hippocampus,the left caudate nucleus,the left lenticular globus pallidus,the left superior temporal gyrus node increased,and these abnormal nodes were basically Belongs to default network(DMN)Conclusion: 1.PSD patients still had small world properties,but the small world property damage,trend of random network topology change;2.nodes topological disorder of the salience network(SN),the default network(DMN),the central executive control network(CEN)can easily lead to the occurrence of PSD.