Research On Voxel-Based Whole Brain Functional Connectivity:Resting-State Functional Connectivity Strength Characteristics Of Amnestic Mild Cognitive Impairment

Background Amnestic mild cognitive impairment(a MCI) is a high risk factor for progressing to Alzheimer’s disease(AD). Previous studies have shown functional connectivity abnormalities in a MCI. However, there are few studies exploring voxelbased whole brain functional connectivity characteristics. This study was designed to explore a new index reflecting the topological information of brain functional network in patients at high risk for AD dementia through investigating resting-state functional connectivity strength(FCS) characteristics of a MCI patients.Materials and methods This study involved 36 a MCI patients and 35 age, genderand years of education-matched healthy controls(HC). The resting-state functionalMRI(rs-f MRI) data of all participants were acquired and preprocessed. Individualwhole brain functional connectivity was constructed for each participant and for agiven voxel, functional connectivity strength(FCS) was computed as the sum of theconnections between this voxel and all others. Using two-sample t test to comparegroup differences in age, years of education, each neuropsychological score andmedial temporal atrophy(MTA). In addition, using Chi-squared test to comparegroup differences in gender. Group differences in FCS value were analysed bygeneral linear model. Then correlation analysis was used to evaluate the relationshipsbetween the FCS values of the brain regions with significant group differences andclinical scores in a MCI patients.Results Regions with higher FCS value were mainly located in the association cortices such as the inferior parietal lobe(IPL), superior occipital gyrus(SOG), dorsolateral prefrontal cortex(DLPFC), posterior cingulate cortex(PCC), etc. both in a MCI and HC. However, FCS values of these brain regions were significantly lower in a MCI patients than those in HC. In comparison to HC, a MCI patients had significantly increased FCS value predominantly in bilateral medial temporal lobe(MTL) and orbitofrontal cortices, while decreased FCS value mainly in bilateral occipital cortices. Area under the curve(AUC) analysis showed that the sensitivity of distinguishing a MCI patients from HC was near eighty percent. Finally, correlation analysis under the uncorrected condition showed that FCS value of the right hippcampus was positively correlated with the auditory verbal learning test-delayed recall(AVLT-D) and AVLT-recognition(AVLT-R) scores(R=0.516; R=0.451 respectively) and similar correlation between left middle temporopolar and AVLT-R score(R=0.572).Conclusion The present study suggests that a MCI has both decreased and increased functional connectivity in some brain regions, meaning that disruption and compensation may be simultaneous. For a MCI patients, increased FCS in medial temporal regions may compensate the cognitive function to some extent, which may serve as a potential biomarker of prodromal AD.