Stratifying Heterogeneous Cognitive Phenotypes in Schizophrenia Through Functional Connectivity Signatures Derived from Resting-State fMRI

Abstract

Schizophrenia is a complex neuropsychiatric disorder characterized by substantial heterogeneity in cognitive functioning. Stratifying cognitive phenotypes in schizophrenia based on functional connectivity (FC) signatures derived from resting-state functional magnetic resonance imaging (rs-fMRI) offers a promising approach to understanding this heterogeneity. This paper explores how rs-fMRI data can be used to identify distinct connectivity patterns associated with different cognitive profiles in schizophrenia. A comprehensive literature review is presented, focusing on studies before 2021 that have examined rs-fMRI and cognitive stratification. Key findings suggest that aberrant connectivity in frontoparietal, default mode, and salience networks underlies cognitive dysfunction in schizophrenia. We present original data analysis and visualizations to support these findings, highlighting how machine learning models can further refine phenotype classification. Finally, we discuss the potential clinical implications of this stratification approach for personalized treatment strategies.