Full correlation matrix analysis of fMRI data

Functional brain imaging produces huge amounts of data, of which only a fraction are analyzed. Existing univariate and multivariate analyses of brain activity ignore interactions between regions, and analyses of interactions (functional connectivity) are typically biased toward regions of interest chosen based on their activity profile. This technical report provides a provisional description of an unbiased technique for functional connectivity, full correlation matrix analysis (FCMA). This technique calculates and analyzes all pairwise relationships between voxels over multiple time windows by leveraging advances in parallel computing and machine learning. FCMA enables the identification of neural mechanisms that support cognitive processes but may be invisible to activity-based methods.