Pyrcca: regularized kernel canonical correlation analysis in Python and its applications to neuroimaging

TL;DR

Pyrcca is an open-source Python tool that implements regularized kernel canonical correlation analysis, enabling neuroimaging researchers to identify shared functional patterns across subjects and predict responses to new stimuli.

Contribution

This paper introduces Pyrcca, a versatile Python module for kernel CCA with regularization, tailored for neuroimaging data analysis and cross-subject functional pattern discovery.

Findings

Identified common functional response patterns across subjects.

Successfully predicted responses to novel stimuli using CCA-derived patterns.

Demonstrated the utility of Pyrcca in neuroimaging applications.

Abstract

Canonical correlation analysis (CCA) is a valuable method for interpreting cross-covariance across related datasets of different dimensionality. There are many potential applications of CCA to neuroimaging data analysis. For instance, CCA can be used for finding functional similarities across fMRI datasets collected from multiple subjects without resampling individual datasets to a template anatomy. In this paper, we introduce Pyrcca, an open-source Python module for executing CCA between two or more datasets. Pyrcca can be used to implement CCA with or without regularization, and with or without linear or a Gaussian kernelization of the datasets. We demonstrate an application of CCA implemented with Pyrcca to neuroimaging data analysis. We use CCA to find a data-driven set of functional response patterns that are similar across individual subjects in a natural movie experiment. We then demonstrate how this set of response patterns discovered by CCA can be used to accurately predict subject responses to novel natural movie stimuli.