I ran into a normalization issue with one participant in particular. This is illustrated in the attached pdf "book", where I am showing the registration of the subject's anatomical to the subject's EPI, then the registration of the SPM EPI template (top) to the subject's realigned, normalized and smoothed image and finally the registration of the SPM EPI template to the subject's realigned, normalized and smoothed image after stretching it by a factor of 1.1 in both the y and z direction. I am showing three views of each sets.
The final result looks ok but I am worried about this ~10% error with the "unified segmentation approach.
What I did, in terms of Analysis is:
1. Reorient anatomical and functional images separately to the AC
Then, use the preprocessing steps (batch) as outlined in the SPM8 manual (expect I didn't write the realigned images and didn't do slice time correction)
Realign:estimate, Coreg:Estimate, Segment, Normalize:Write, Smooth.
University of Alabama