I have two questions:

1) if I do recon-all on a group of 20 subject, and I am primarily interested in a few regions (In my case - amygdala, insula and anterior cingulate cortex), do I have to edit all the datasets in tkmedit to make sure that these ROIs are "where they ar supposed to be"? 

2) What is the easiest way to convert recon'ed data from FreeSurfer space to original subject space ( and mgz files to nifti files)?

Thank you so much for your help!

Kind regards, Jenny