Analyzing brain structural differences among undergraduates with different grades of self-esteem using multiple...

Area of Science:

• Neuroscience
• Brain Imaging
• Psychology

Background:

• Self-esteem, an individual's self-evaluation, is crucial for mental health and coping abilities.
• Neuroimaging research on the cognitive neural mechanisms of self-esteem is expanding.
• Current methods using brain morphometry and single-layer networks struggle to capture subtle structural differences linked to self-esteem.

Purpose of the Study:

• To develop a novel method for studying brain structural connections associated with self-esteem.
• To address limitations of existing techniques in characterizing subtle structural differences.
• To enhance the accuracy of identifying brain structural variations related to self-esteem.

Main Methods:

• Proposed a multiple anatomical brain network approach using multi-resolution region of interest (ROI) templates.
• Extracted ROI and hierarchical brain network features from structural MRI.
• Employed feature selection (t-test, mRMR, SVM-RFE) and multi-kernel SVM for robust analysis.

Main Results:

• The proposed multiple anatomical brain network method achieved a classification accuracy of 97.26%.
• This represents a significant improvement over single-layer brain network approaches.

Conclusions:

• The developed method offers a new perspective for analyzing brain structural differences in self-esteem.
• This approach holds potential guiding significance for research on brain cognitive activity and disease diagnosis.