I encountered a weird problem with the "new segment" in spm8. When I used it to segment brains with severe atrophy, it seems that "new segment" filled in the atrophied region with additional gray matter. On the other hand, "segment" (the old version) could represent the actual atrophy just fine. I wonder whether the new segment tried too hard to match to the normal template, which caused distortion in injured brains? Or there is some setting with "new segment" I could adjust to avoid this problem.
Thanks for your advice.