Association between abnormal brain functional connectivity in children and psychopathology: A study based on graph...

Area of Science:

• Neuroscience
• Psychiatry
• Computational Biology

Background:

• Incorporating biological measures into mental health disorder classification remains a challenge.
• Mental health disorders often impact brain development and neural connectivity.

Purpose of the Study:

• To propose a novel method for assessing brain networks using graph theory and machine learning.
• To evaluate the method's ability to identify typical versus atypical brain networks and predict psychopathology levels.

Main Methods:

• Applied a novel approach combining eigenvector centrality (a graph theory measure) and a one-class support vector machine (OC-SVM).
• Utilized resting-state functional magnetic resonance imaging (fMRI) data from 622 children and adolescents.
• Extracted eigenvector centrality (EVC) from positive and negative task networks to train the OC-SVM for network classification.

Main Results:

• Atypical brain network organization was significantly associated with higher levels of psychopathology (p < 0.001).
• The typical group showed greater EVC in bilateral posterior cingulate and temporal cortices.
• Significant decreases in EVC were observed in the left temporal pole for the atypical group.

Conclusions:

• The combination of graph theory and OC-SVM is a promising method for characterizing neurodevelopment.
• This approach may offer insights into the deviations underlying mental health disorders.
• The findings highlight the potential of integrating neuroimaging and computational methods in psychiatric research.