BrainStat: A toolbox for brain-wide statistics and multimodal feature associations

Area of Science:

• Neuroscience
• Bioinformatics
• Computational Biology

Background:

• Neuroimaging analysis increasingly requires integrating diverse data types beyond statistics.
• Associations with gene expression, histology, and functional/cognitive architectures are crucial for comprehensive brain analysis.

Purpose of the Study:

• Introduce BrainStat, a novel toolbox for neuroimaging data analysis.
• Facilitate univariate and multivariate linear modeling in volumetric and surface-based brain imaging.
• Enable multidomain feature association with gene expression, histology, fMRI meta-analysis, and resting-state fMRI motifs.

Main Methods:

• BrainStat implements linear models for volumetric and surface-based neuroimaging data.
• It performs multidomain feature association with spatial maps of gene expression, histology, and fMRI data.
• The toolbox is available in Python and MATLAB.

Main Results:

• BrainStat streamlines the integration of statistical analysis and feature association.
• It accelerates cross-modal research in neuroimaging.
• The toolbox supports common surface templates for consistent analysis.

Conclusions:

• BrainStat offers a comprehensive solution for analyzing and interpreting complex neuroimaging datasets.
• The toolbox facilitates interdisciplinary research by bridging statistical and biological data.
• Its open availability and documentation support widespread adoption and further development.