Analysis Of Brain Image Interaction Effect And Surface Characteristics Based On ANOVA Model

Exploring the interactive influence of multiple factors on brain structure image and brain metabolism image has always been one of the most important means in brain imaging research,and Two-way ANOVA has also been widely used in brain imaging research.For brain imaging studies,previous studies mainly focused on the average metabolic intensity of brain regions,and used the graph theory characters to analyze the inter-group differences of brain networks,while the analysis of complex inter-group interaction effects caused by multiple factors and voxel-level brain metabolic effects was rarely concerned by researchers.Recent studies showed that the APOE gene and the CLU gene are closely related to human cognitive ability,structural abnormalities in specific brain regions,and human Alzheimer 's disease.As APOE and CLU are the two risk genes related to Alzheimer 's disease,it is particularly important for us to explore whether the two risk genes have an interactive effect on the brain structure of the subjects.Based on the above background,the main work of this paper is divided into the following two parts:In the first part of this paper,in order to investigate the effects of gender and cancer factors on cerebral brain metabolism,a two-way ANOVA was performed to analyze the effects of gender and cancer on cerebral brain metabolism.We recruited 49 pretreatment cancer patients with a diagnosis time of less than one year and 48 Normal controls whose gender and age could be matched,and performed two-way ANOVA in voxel-level on their brain metabolism images.Next,we used the clusters of main effects,interaction effects as ROI(Region of interest)to conduct the correlation analysis with depression scale of cancer patients.Our results showed that the cancer main effect mainly existed in the superior frontal gyrus and parietal lobe,the gender main effect was found in the inferior temporal gyrus and posterior cerebellar lobe.The interaction effect of the two factors mainly existed in the superior frontal gyrus of the subjects.Ourresults suggested that in the whole brain metabolism of cancer patients,some areas are mainly affected by cancer factor,some areas are mainly affected by gender factor,and some areas are jointly affected by two factors.The results of the correlation analysis in this paper showed that there were voxel clusters significantly correlated with the BDI score in the ROI extracted from the anova experiment,which mainly located in the superior frontal gyrus,superior superior gyrus,cuneus and inferior temporal gyrus.These results showed that there were clusters in the brains whose metabolic intensity was affected by both cancer and gender factors,and the metabolic intensity of these clusters was significantly correlated with the depressive symptoms of the subjects.The results above suggested that the brain metabolism of cancer patients and the depressive symptoms are affected by gender and cancer factors.And these brain regions should be concerned in clinic.In the second part of this paper,we conducted a two-way ANOVA on the effect of APOE and CLU risk genes on specific brain structures.284 subjects with genetic data were chosen from the ADNI database,and their hippocampal and amygdala volumes,as well as the surface characteristics were calculated.Then,we used the ANOVA to explore whether there was a significant interaction between the two risk genes and the structure indexes of the hippocampus and amygdala.The results showed that the main effect of CLU risk gene is existed on the amygdala volume.In the aspect of surface characteristics,the CLU main effects were significant on the surface characteristics of the subjects' hippocampus and amygdala.The APOE gene only had significant main effects on the surface characteristics of the subjects' hippocampus.The interaction effects of the two risk genes also existed in the surface characteristics of the subjects' hippocampus.Our results suggested that the surface features of subjects will be changed before the volume of the hippocampus and amygdala of subjects.The surface features of the hippocampus and amygdala of subjects should be paid more attention in clinical practice as an early biological marker of Alzheimer's disease.