A Study Of Brain Functional Networks In Patients With Alzheimer’s Disease And Mild Cognitive Impairment Based On Chinese Brain Atlas

Objectives: The main objectives of this study were exploring the changes of the topological attributes of brain functional networks including cerebellum in patients with Alzheimer’s disease(AD)and mild cognitive impairment(MCI)based on Chinese brain atlas,and their relationship with the cognitive function and activities of daily living;exploring the expression of brain plasticity in AD and MCI patients’ brain functional network topology;and obtaining imaging biomarkers that can be used in diagnosis of diseases.Contents and Methods: Subjects were divided into AD group(n = 24),MCI group(n = 27),Healthy control(HC)group(n = 33).All subjects were Han Chinese and dextromanual.Resting-state functional magnetic resonance(rs-f MRI)data of all subjects was collected.Data preprocessing and brain functional networks construction were performed based on Chinese specific atlas Chinese2020 to improve the accuracy of segmentation,registration and network nodes definition.Graph theory analysis method was used to calculate multiple brain functional network topological properties.Global network metrics included clustering coefficient,characteristic path length,local efficiency,global efficiency,modularity,assortativity and synchronization.Nodal network metrics included nodal clustering coefficient,nodal shortest path length,nodal efficiency,nodal local efficiency,nodal degree centrality and nodal betweenness centrality.Several neuropsychological scales were also used to evaluate the cognitive function and activities of daily living of the subjects,including Mini-Mental State Examination(MMSE),Montreal Cognitive Assessment(Mo CA),Clinical Dementia Rating(CDR),Activities of Daily Living(ADL),Hamilton Depression Scale(HAMD),Hachinski Ischemic Score(HIS),Neuropsychiatric Inventory(NPI)and Alzheimer’s Disease Assessment Scale-Cognitive Subscale(ADAS-Cog).In the first part of the study,all the brain functional network metrics of the three groups were analyzed using Analysis of Variance(ANOVA)with false discovery rate(FDR)corrected and post hoc pairwise comparisons.In the second part of the study,the correlations between the graph metrics with significant differences after ANOVA and the MMSE,Mo CA,ADL and CDR scale scores were analyzed(FDR corrected),and the correlations between the area under curve(AUC)of the network sparsity curve for all graph metrics(including all global metrics and nodal metrics)of all subjects and the MMSE,Mo CA,ADL and CDR scale scores were analyzed.Results: Assortativity of the whole brain,nodal degree centrality of the left middle frontal gyrus,nodal degree centrality of the left medial superior frontal gyrus,nodal degree centrality of the right superior cerebellum,nodal clustering coefficient of the right superior cerebellum,nodal efficiency of the left middle frontal gyrus,nodal local efficiency of the left superior cerebellum,nodal local efficiency of the right superior cerebellum,and nodal local efficiency of the vermis were significantly different among the three groups.In post hoc pairwise comparisons,compared with healthy controls,patients with AD showed significantly higher degree centrality in the left middle frontal gyrus and left medial superior frontal gyrus,but lower degree centrality in the right superior cerebellum,as well as higher nodal efficiency in the left middle frontal gyrus;compared with healthy controls,patients with MCI showed significantly higher degree centrality in the left middle frontal gyrus,lower degree centrality in the right superior cerebellum,lower nodal clustering coefficient in the right superior cerebellum,higher nodal efficiency in the left middle frontal gyrus,and lower nodal local efficiency in the left,right superior cerebellum and vermis;compared with patients with MCI,patients with AD showed significantly higher degree centrality in the left middle frontal gyrus and left medial superior frontal gyrus,higher nodal clustering coefficient in the right superior cerebellum,higher nodal local efficiency in the left middle frontal gyrus,and lower nodal local efficiency in the left,right superior cerebellum and vermis.Several global and nodal metrics demonstrated significant positive or negative correlations with MMSE,Mo CA,ADL,and CDR scores.During the progressions of MCI and AD,the nodes with reduced communication capability were mainly concentrated in the temporal lobe(about 66%),and a few were distributed in the limbic system(about 17%)and parietal lobe(about17%).The left temporal lobe nodes accounted for 53% of all the damaged temporal lobe nodes.The left hemisphere nodes accounted for 48% of all damaged nodes.The nodes with restored communication capability were mainly concentrated in the frontallobe(about 76%)and a few were distributed in the cerebellum(about 10%),parietal lobe(about 6%),limbic system(about 4%)and temporal lobe(about 4%).The right frontal lobe accounted for 50% of the functional remodeling frontal nodes.But accounted for 60% of the functional remodeling nodes related to activities of daily living.Right hemisphere nodes accounted for 57% of all the nodes with functional plasticity.Conclusion:As a brain functional network graph analysis on AD and MCI based on the population-specific Chinese brain atlas,this study may aid future studies on graph analysis of brain functional networks in neuroimaging based on specific brain atlases for different ethnicities.MCI patients may have unique alternations of functional network topological attributes in cerebellar regions.Damage and remodeling of brain functional networks were common in progressions of MCI and AD.Disease attacks on hub nodes of the left and right hemispheres were almost equal and did not differentiate for dominant hemisphere.However,the non-dominant hemisphere played a more important role in the brain function remodeling in response to AD and MCI attacks,especially in the function remodeling related to activities of daily living.Moreover,in the progression of disease,the damage of temporal lobe functional network was more serious.And as the center of many advanced activities,frontal lobe took more responsibilities for functional compensation and remodeling.Graph metrics with significant differences among AD,MCI and HC groups may become imaging biomarkers used for diagnoses of AD and MCI.