Network biomarkers of schizophrenia by graph theoretical investigations of Brain Functional Networks

TL;DR

This study identifies key network markers in brain functional networks derived from fMRI data that distinguish schizophrenia patients from healthy individuals, offering potential for improved diagnosis and understanding of the disorder.

Contribution

It introduces a comprehensive analysis of topological features in weighted and binary brain networks to find biomarkers specific to schizophrenia.

Findings

Features related to modularity, betweenness, assortativity, and edge density are key markers.

Weighted network features outperform binary features in disease classification.

Network markers may aid in clinical diagnosis and early detection of schizophrenia.

Abstract

Brain Functional Networks (BFNs), graph theoretical models of brain activity data, provide a systems perspective of complex functional connectivity within the brain. Neurological disorders are known to have basis in abnormal functional activities, that could be captured in terms of network markers. Schizophrenia is a pathological condition characterized with altered brain functional state. We created weighted and binary BFN models of schizophrenia patients as well as healthy subjects starting from fMRI data in an effort to search for network biomarkers of the disease. We investigated 45 topological features of BFNs and their higher order combinations (2 to 8). We find that network features embodying modularity, betweenness, assortativity and edge density emerge as key markers of schizophrenia. Also, features derived from weighted BFNs were observed to be more effective in disease classification as compared to those from binary BFNs. These topological markers may provide insights into mechanisms of functional activity underlying disease phenotype and could further be used for designing algorithms for clinical diagnosis of schizophrenia as well as its early detection.