Hello,

I am doing a seed-based analysis based on ROIs previously defined (in MNI space) for specific regions of S1. For several subjects, these ROIs fall outside, or only partially inside, the brain following direct normalization. No fieldmaps or topups for distortion correction are available, but I tried a few different variations of using flirt and fnirt to use the anatomical images for each subject to register the functional images indirectly to MNI space. However this did not yield much better results, perhaps because I was using the wc0* images? (As far as I can tell CONN does not create skull stripped images in the native anatomical space). If anyone has any suggestions please let me know. I've attached a screenshot to give an example of the seed falling partially outside of the brain.

Best,
- Harris