Decoding gender dimorphism of the human brain using multimodal anatomical and diffusion MRI data.

Area of Science:

• Neuroimaging
• Neuroscience
• Medical Imaging Analysis

Background:

• Human brain architecture exhibits sex-specific differences, known as brain dimorphism.
• Previous studies primarily used unimodal MRI, limiting the scope of observed gender-related brain variations.
• Morphological information in MRI is contrast-dependent, necessitating multimodal approaches for comprehensive analysis.

Purpose of the Study:

• To develop and validate a novel multimodal MRI approach for classifying brain dimorphism.
• To investigate the combined utility of T(1)-, T(2)-, and diffusion-weighted imaging for gender-specific brain architecture analysis.
• To identify specific brain regions and networks most discriminative of sex differences.

Main Methods:

• Analysis of multimodal MRI data (T(1)-, T(2)-, diffusion-weighted) from 121 subjects.
• Application of a linear support vector machine classifier with mass-univariate feature selection.
• Comparison of classification accuracy between unimodal and multimodal approaches.

Main Results:

• Multimodal MRI classification achieved 96% accuracy, significantly outperforming unimodal methods (83%-88%).
• Identified key gender disparities in gray matter volume and white matter microstructure.
• Discrepancies were concentrated in brain networks associated with social cognition, reward processing, decision-making, and visuospatial skills.

Conclusions:

• Multimodal MRI is superior to unimodal approaches for detecting sex-based brain differences.
• The study refines understanding of brain dimorphism, highlighting specific network involvements.
• Findings contribute to a more nuanced view of gender-related variations in brain structure and function.