Intrinsic Discriminant Analysis of Functional Connectivity for Multiclass Classification

Abstract

Major depression and schizophrenia are two of the most serious diagnosed psychiatric disorders that share some similar behavioral symptoms. Whether these similar behavioral symptoms underlie any convergent psychiatric pathological mechanisms is to date unclear. In this chapter, the whole-brain resting-state functional magnetic resonance imaging of major depression and schizophrenia was investigated by using intrinsic discriminant analysis, which is a supervised linear dimensionality reduction method maximizing the inter-class difference while minimizing the intra-class difference. Thirty-two schizophrenic patients, 19 major depressive disorders, and 38 healthy controls underwent the resting-state functional magnetic resonance imaging scanning. Support vector machine in conjunction with intrinsic discriminant analysis was used to solve the multi-classification problem, resulting in a correct classification rate of 80.9% via leave-one-out cross-validation. It was revealed that the depression and schizophrenia groups both showed altered functional connections related to the medial prefrontal cortex, anterior cingulate cortex, thalamus, hippocampus, and cerebellum. However, the prefrontal cortex, amygdala, and temporal poles were found to be different affected between major depression and schizophrenia. This preliminary study suggests that altered connections within or across the default mode network and cerebellum might account for the common behavioral symptoms between major depression and schizophrenia. In addition, connections related to the prefrontal cortex and affective network showed promised as biomarkers in discrimination between the two disorders.