Combining graph and machine learning methods to analyze differences in functional connectivity across sex.

Area of Science:

• Neuroscience
• Machine Learning
• Graph Theory

Background:

• Resting-state functional magnetic resonance imaging (fMRI) allows for the study of intrinsic brain function.
• Investigating gender-associated differences in brain connectivity is crucial for understanding neurodevelopment and neurological disorders.

Purpose of the Study:

• To employ machine learning and graph theory to identify gender-specific differences in resting-state brain network connectivity.
• To detect discriminative network connections between male and female subjects.

Main Methods:

• Utilized resting-state fMRI data from the Connectome Project.
• Applied ensemble learning methods, including Random Forest and Least Angle Shrinkage and Selection Operator (Lasso) regression, for classification.
• Employed permutation testing to validate classification accuracy and feature selection significance.

Main Results:

• Identified specific brain network connections (edges) that significantly differentiate between male and female subjects.
• Machine learning models achieved significant classification accuracy in distinguishing between genders based on brain connectivity patterns.
• Feature selection highlighted critical nodes and gender-discriminative edges.

Conclusions:

• Gender differences in brain function are associated with sexually dimorphic regional connectivity.
• Specific critical brain nodes are connected via gender-discriminative edges, influencing functional network organization.
• This approach provides a robust method for identifying neurobiological correlates of gender differences.