A likely naive question…

I am running my analysis through FSL for specific regions of interest of a one-sample t-test at the group-level. Originally, I had planned to include group as a variable for the “online” FEAT analysis with an unpaired t-test (non-parametric TFCE correction). But now I would like to include other variables as predictors in a model from other tasks. So, I would like to enter brain activation into an offline model (in R or JASP, for example) with these other variables.

My question is what kind of values/output would I need from FSL for such a model, and how would I extract this as an output?