Classification of schizophrenia and bipolar patients using static and time-varying resting-state FMRI brain connectivity

Recently, there is a growing interest in designing objective prognostic/diagnostic tools based on neuroimaging and other data that display high accuracy and robustness. Small training subjects and very large amount of high dimensional data make it a challenging task to design robust and accurate classifiers for heterogeneous disorders such as schizophrenia. Majority of previous works have focused on classification of schizophrenia from healthy controls while automatic differential diagnosis of schizophrenia from bipolar disorder has been rarely investigated. In this work, we propose a framework for automatic classification of schizophrenia, bipolar and healthy control subjects based on static and dynamic functional network connectivity (FNC) features. Our results show that disrupted functional integration in schizophrenia and bipolar patients as captured by FNC analysis reveal powerful information for automatic discriminative analysis.