I had a couple of questions. I currently am using custom ROIs that exist in each individual's coordinate space, and I know that during the preprocessing steps in CONN, the program does a transform from individual subject space to MNI space on the brain scans themselves, but does the subject space to MNI space transformation happen to the custom ROIs as well or are they just directly imported and analyzed as if they were already in MNI space?

If there is no ROI transformation to MNI space, is the transformation matrix saved as a variable and exported in one of the mat files? If not, where can I find the transformation in the code? I would like to save the transformation matrix during the analysis and then use that on the ROIs for each subject to get them into the same space.

If it makes a difference, I am using the batch functions for all of the preprocessing and first level analyses.