Dear SPMers,
I have fMRI data and T1 weighted anatomical scans from 80 subjects, recorded with a 3T scanner using a 32-channel headcoil. After preprocessing the data with SPM8 (slice-timing correction, realignment, co-registration) co-registration of the anatomical onto the mean EPI gave suboptimal results in 4 subjects. I followed the advice of John Ashburner of using a bias-corrected, skull-stripped version of the anatomical for co-registration as outlined here:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0906&L=SPM&P=R49092
This worked perfectly fine and I could correctly register the anatomical to the mean EPI in those subjects. But after finishing my preprocessing pipeline (New Segment: Dartel imported rc1 & rc2 images from the anatomical scans; feeding rc1&rc2 images into the DARTEL toolbox, creating a group template and individual flow fields and subsequently using them to normalize the functional images) normalization totally failed for 1 subject. It was one of the subjects which I co-registered using the bias-corrected and skull-stripped anatomical. Normalization worked fine for all the other subjects.
I looked at the rc1 & rc2 image of that subject and they look very strange (see attachment). I repeated the procedure (co-registration as outlined above + segmentation) several times but i always get the same results (even if I turn up the bias regularisation in New Segment to "very heavy").
Can anyone tell me what went wrong and what I can do about it?
Any help would be very much appreciated!
Thanks,
Christian
|