Early Diagnosis And Intervention Of Mild Cognitive Impairment Based On Brain Function Network

Alzheimer’s disease(AD)is the most common type of Senile Dementia,its pathogenesis is unknown and there is no effective treatment.Mild cognitive impairment(MCI)is the early manifestation of AD,and the effective diagnosis and intervention of MCI can reduce the transformation to dementia.The results show that the brain function connection of MCI patients has been abnormal.Based on the analysis of brain function connection characteristics,this paper explores the intervention effect of MCI diagnosis index and neural feedback training,hoping to provide help for the early diagnosis and auxiliary intervention of MCI.Based on Granger dynamic causal analysis,the direct transform function(DTF)is used as the index of brain function connection,and the absolute threshold method and relative threshold method are used to determine the optimal threshold of brain function connection,and the brain function network is constructed.The characteristic parameters of brain function network are analyzed further.Especially,the beta band and gamma band with significant difference are the main analysis frequency bands.The characteristics of MCI brain function network connection are analyzed from the graph theory and information transmission direction of brain function network.Based on the extraction of traditional graph theory features such as node degree and clustering coefficient,a new feature parameter of brain function network called efficiency density,is proposed,which can better reflect the global information transmission ability of brain function network from the two aspects of node aggregation and shortest feature path.And based on SVM diagnosis model,the results of MCI diagnosis are verified.The results show that: the average diagnostic accuracy,sensitivity and specificity of MCI group and normal control group(NC)are86.25%,86.84% and 85.71%,respectively,which indicates that the diagnosis of MCI can be effectively realized based on the brain functional network features with efficiency density as the core.Furthermore,for MCI patients,two cycles of neural feedback training were conducted to compare the effects of neural feedback training on brain functional connection.The training effect was evaluated by using Montreal Cognitive Assessment Scale(Mo CA),cognitive ability screening scale(CASI),power spectral density and clinical diagnosis of physicians as support.The results show that the DTF matrix connection increases significantly after training,and the number of channels with weak causal relationship decreases;The causal connection of brain function network was enhanced at the optimal threshold,and the level of the first cycle was higher than that of the second cycle.The graph theory features under the optimal threshold were higher than those before training except clustering coefficient(P<0.05).The results of psycho neuro scale and power spectrum density analysis showed that MOCA and CASI scores increased after training,and there was significant difference(P<0.05).The power spectral density of frontal lobe in beta band gradually increased,and that of right temporal lobe decreased first and then increased in beta and gamma bands.The results show that neural feedback training can improve the connection state of brain function network in MCI patients to some extent,and can be used as an auxiliary intervention method for MCI.