A New Approach for Predicting Brain Functional Connectivity: A Cross-Modality DTI-fMRI Study.

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Brain Connectivity Research

Background:

• Diffusion tensor imaging (DTI) and functional magnetic resonance imaging (fMRI) offer distinct views of brain connectivity: DTI maps structural connectivity (SC), while fMRI measures functional connectivity (FC).
• Integrating DTI and fMRI data can enhance understanding of the brain's structural-functional coupling, the relationship between physical brain structure and its activity patterns.
• Existing methods for integrating DTI and fMRI are limited, necessitating novel approaches to accurately predict functional connections from structural data.

Purpose of the Study:

• To develop and validate a novel integrated DTI-fMRI approach for predicting brain functional connections based on underlying structural pathways.
• To investigate the influence of structural properties (e.g., SC strength, path length, physical distance) on functional connectivity (FC) prediction.
• To assess the stability of structure-function coupling across adulthood and between sexes, and identify age-related variations in predictive accuracy.

Main Methods:

• Developed an asymmetric DTI-driven fMRI model using a linear combination of DTI-derived features to predict FC.
• Trained the model on a cohort of healthy young adults (n=12) using 4-fold cross-validation.
• Validated the model on an independent cohort of healthy adults (n=14) across a wider age range, assessing age and sex effects on predictive performance.

Main Results:

• Cross-validation demonstrated that SC strength, path length, and physical distance significantly predicted FC.
• The integrated model showed whole-brain predictive similarity between actual and estimated FC that was independent of age and sex.
• Age-related differences were observed in link-level reconstruction errors, particularly in frontal brain connections, indicating localized variations in structure-function coupling.

Conclusions:

• The DTI-fMRI integration approach successfully predicts functional connectivity from structural data, highlighting the predictive power of structural network properties.
• Brain structure-function coupling appears robust and stable throughout adulthood, irrespective of sex.
• Localized age-related differences in structure-function coupling exist, primarily affecting frontal connections, suggesting potential targets for future research in aging and neurological conditions.