Multivariate Pattern Analysis Of Schizophrenia By Using Resting-state Fmri

Schizophrenia is a severe mental illness, and it is generally believed that patients with schizophrenia have abnormal brain regions or functional connectivity.This paper investigated the functional connectivity of schizophrenia patients and healthy control subjects during resting state by a multi-voxel pattern analysis, which aimed to identify the abnormal functional connectivity network and brain regions.The whole-brain, resting-state, fMRI was performed on22cases of patients with schizophrenia and22healthy control subjects. Patients were evaluated based on a patient version of the Structured Clinical Interview for DSM-IV (Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition), and the two groups were matched in terms of gender, age and education. After preprocessing the fMRI data, we extracted the functional connectivity as the feature, and then chose Locally linear embedding, Linear Discriminant Analysis and Principle Component Analysis to reduce the dimensionality of the selected features, respectively, and the final step was classifying the data by using C-means clustering and Support vector machine, respectively. After the comparative analysis of the classification results obtained by different methods, the weighted functional connectivity and brain regions were got by reconstruction arithmetic to extract the highly discriminative functional connectivity information for the results of the optimal classification rate.The results showed that the method integrated principle component analysis and support vector machine with linear kernel achieved the best classification rate, and the classification rate was93.2%. And most of the connectivity features located in the visual cortex network, the default mode network and sensory-motor network. Moreover, in the reconstruction algorithm, the fusiform gyrus exhibited the greatest weight.This study demonstrates that schizophrenia patients can be successfully differentiated from healthy controls by using the multi-voxel pattern analysis, and the fusiform gyrus may play an important role in the physiological symptoms manifested by schizophrenic patients. The brain region of great weight may be the problematic region in information exchange of schizophrenic patients. Therefore, the results of this paper will provide insights into the identification of potentially effective biomarkers for the clinical diagnosis of schizophrenia.