Dear Kyle,
Sometimes the segmentation may produce undesirable results when the
structurals are considerably misaligned to the priors. A way to deal with
this is:
1) Coregister (estimate only) your structural to the T1 template (the T1.nii
file that comes with SPM);
2) Segment your coregistered structural (save in the native space);
3) If you need the segmented image aligned with the functionals, then now
you can coregister again (estimate only) your original structural to the
mean functional and select the newly created c1* image as "Other images".
The "Estimate Only" option when coregistering is to prevent unnecessary
resampling artifacts, so only the affine matrix in the header will be changed.
Hope this helps!
Best regards,
Anderson
|