A MATLAB toolbox for multivariate analysis of brain networks.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Graph Theory Applications

Background:

• Complex brain networks underpin cognitive functions.
• Altered network topology correlates with various abnormalities.
• Existing methods lack robust statistical inference for population studies.

Purpose of the Study:

• To develop a freely available MATLAB toolbox for statistical inference in brain network analysis.
• To bridge the gap between brain network analysis and statistical modeling.
• To enable assessment of covariates on brain network properties.

Main Methods:

• Developed a MATLAB toolbox with a graphical user interface.
• Implemented a mixed-effects multivariate regression framework.
• Incorporated a clustering-based data reduction method for efficiency.
• Supported analysis of fMRI, EEG, and DTI data.
• Utilized SAS, R, or Python for statistical modeling.

Main Results:

• The toolbox facilitates statistical inference on brain network differences across populations.
• It allows assessment of covariates (age, disease) on global and local brain network properties.
• Confounding variables like sex are controlled for.
• A data reduction method improves model convergence and reduces computation time.

Conclusions:

• The developed toolbox provides a powerful, accessible platform for advanced brain network analysis.
• It enables robust statistical assessment of factors influencing brain connectivity.
• This tool is valuable for diverse neuroimaging studies investigating brain structure and function.