A Transdiagnostic Biotype Detection Method for Schizophrenia and Autism Spectrum Disorder Based on Graph Kernel

Area of Science:

• Neuroscience
• Psychiatry
• Computational Biology

Background:

• Psychiatric diagnoses relying solely on clinical symptoms can be imprecise.
• Schizophrenia (SZ) and Autism Spectrum Disorder (ASD) share overlapping symptoms, necessitating objective diagnostic markers.
• Brain functional connectivity (FC) offers a potential neuroimaging biomarker for psychiatric disorders.

Purpose of the Study:

• To develop and validate a novel method for identifying transdiagnostic biotypes in psychiatric disorders.
• To leverage graph theory and functional connectivity (FC) for improved biotype detection.
• To explore potential shared neurobiological underpinnings between Schizophrenia and Autism Spectrum Disorder.

Main Methods:

• A graph kernel-based clustering method was developed to analyze whole-brain functional connectivity (FC) from functional magnetic resonance imaging (fMRI) data.
• Frequent subnetworks were identified, and graph kernel similarity was computed to cluster subjects.
• The method was applied to fMRI data from 137 SZ and 150 ASD subjects.

Main Results:

• The proposed graph kernel-based clustering method successfully identified meaningful biotypes within the combined SZ and ASD cohort.
• Significant differences in functional connectivity patterns were observed between the identified biotypes.
• The results suggest distinct neurobiological profiles underlying these transdiagnostic biotypes.

Conclusions:

• The graph kernel-based clustering approach is a promising tool for detecting transdiagnostic biotypes in psychiatric research.
• This method enhances the utility of functional connectivity (FC) in psychiatric diagnosis and subtyping.
• Identifying objective neuroimaging-based biotypes can lead to more accurate and personalized psychiatric care.