Statistical Study Of FMRI Data Features For Aggressive Schizophrenia Patients

Schizophrenia is a serious mental health problem, which may result in violentoffending as compared with the general population. But the causes of violent behaviorpathology of the patients are poorly understood. Functional magnetic resonanceimaging (fMRI) is a noninvasive technique that has advantages in spatial andtemporal resolution, and is becoming an important tool for the research of brainscience. In this paper, the resting state fMRI data of male patients with schizophreniawere analyzed to find out the differences of functional activity in default modenetwork among patients with and without aggression. This research may contribute toclinical diagnosis and pathological study of aggressive schizophrenia patients, and isof great theoretical and paractical importance. The study methods and mainconclusions of this paper are as follows:(1) Regional homogeneity was applied to find the difference at local brain level.First, the Kendall concorde of coefficient was used to calculate the time seriescorrelation of a voxel with its neighbour voxels. It was found that aggressiveschizophrenia patients show local function abnormal activity in the temporal lobe andfrontal brain regions. ROC curve analysis method was used for further analysis ofthese brain regions and a potential index was proposed for judgement.(2) Independent component analysis was applied to investigate on the subsystemfunctional network level to explore the connection strength in the default modenetwork. With the independent component analysis method, the spatial independentcomponent was obtained, and the default mode network space composition of eachpatient was obtained by fit goodness analysis. Finally, the default mode networkinternal function connection intensity of two groups were analyzed. It was found thatthe frontal lobe of patients with aggression has abnormal functional connection withbrain regions in default mode network.(3) Both local brain function and brain interval connection show abnormalities ofthe right frontal lobe. In order to further explore the right frontal lobe in the role ofaggressive behavior, seed point correlation analysis was applied to caluculate theconnection strength of the right frontal lobe with other voxels of the whole brain. Itwas found that aggressive schizophrenia patients have abnormal connection in the right frontal gyrus with other brain regions in default mode network, which may leadto aggressive behaviour of the schizophrenia patients.