Improved fMRI-based Pain Prediction using Bayesian Group-wise Functional Registration

TL;DR

This paper introduces a Bayesian group-wise functional registration method to improve fMRI-based pain prediction by addressing interindividual differences in brain anatomy and functional localization.

Contribution

The study proposes a novel Bayesian registration technique that enhances the alignment of functional brain data across subjects for better predictive modeling.

Findings

Improved alignment of fMRI data across subjects.

Enhanced accuracy in pain prediction models.

Reduction of interindividual variability effects.

Abstract

In recent years, neuroimaging has undergone a paradigm shift, moving away from the traditional brain mapping approach toward developing integrated, multivariate brain models that can predict categories of mental events. However, large interindividual differences in brain anatomy and functional localization after standard anatomical alignment remain a major limitation in performing this analysis, as it leads to feature misalignment across subjects in subsequent predictive models.