Multiplex Community Detection for Subgroup Identification within Functional Connectivity Networks.

Area of Science:

• Neuroscience
• Psychiatry
• Data Science

Background:

• Understanding psychotic disorder heterogeneity is crucial for precision medicine.
• Current methods using functional magnetic resonance imaging (fMRI) often analyze single functional networks, limiting insights.
• A need exists for methods that explore relationships across multiple functional networks.

Purpose of the Study:

• To develop a novel method for identifying patient subgroups in psychotic disorders by integrating cross-functional network information.
• To leverage multiplex network construction and community detection for subgroup analysis.
• To enhance the understanding of psychotic disorder heterogeneity and advance precision medicine.

Main Methods:

• Constructed multiplex networks from functional connectivity networks derived from multi-subject resting-state fMRI data.
• Applied multiplex network-based community detection to identify common and private communities across networks.
• Validated the method using simulations and real-world fMRI data from 464 psychotic patients.

Main Results:

• The proposed method effectively identified clinically relevant subgroups in psychotic patients.
• Identified subgroups showed significant differences in functional areas like the default mode network (DMN) and anterior prefrontal cortex (antPFC).
• Differences in functional areas correlated with distinct clinical scores, aligning with prior research.

Conclusions:

• The novel multiplex network approach successfully uncovers clinically meaningful subgroups in psychotic disorders.
• This method enhances the understanding of individual variability and heterogeneity in psychotic disorders.
• The findings provide a framework for developing tailored treatments and advancing precision medicine in psychiatry.