Estimating functional brain networks by incorporating a modularity prior.

Area of Science:

• Neuroscience
• Network Science
• Medical Imaging

Background:

• Functional brain network analysis is key for understanding brain organization and diagnosing neurological disorders.
• Constructing accurate brain networks from functional magnetic resonance imaging (fMRI) data is challenging due to noise and limited understanding of brain complexity.
• Existing methods struggle to create biologically meaningful and statistically robust brain networks.

Purpose of the Study:

• To develop a novel functional brain network modeling scheme incorporating a modularity prior.
• To formulate network construction as a sparse low-rank graph learning problem.
• To apply the learned brain networks for improved diagnosis of mild cognitive impairment (MCI).

Main Methods:

• Proposed a matrix-regularized network learning framework encoding a modularity prior.
• Formulated the problem as a sparse low-rank graph learning task.
• Solved the model using an efficient optimization algorithm.

Main Results:

• Achieved 89.01% classification accuracy in identifying MCI patients from normal controls.
• Demonstrated superior performance compared to conventional brain network construction methods.
• Identified potential brain network biomarkers contributing to MCI diagnosis.

Conclusions:

• The proposed modularity-informed network learning framework offers a robust method for brain network construction.
• This approach significantly enhances the accuracy of MCI identification.
• The identified biomarkers hold promise for personalized diagnosis of neurological disorders.