Hi all,
I just ran the longitudinal VBM pipeline on my pre- and post-treatment data following the manual ( SPM8 default normalization + default Tissue Probability Map). When I run the flexible factorial design(only one group, no healthy controls) in SPM, I get these wide spread results(FWE corrected at voxel level for Post-treatment GMV > Pre-treatment GMV) throughout the medial surface of brain and the cerebellar region.
I also tried longitudinal VBM in a pre- post- dataset from different treatment. And, the results seem to be very similar, i.e.: too widespread and mostly in the medial region of the brain and cerebellum.
This makes me concerned about the way I performed the longitudinal VBM preprocessing might not be correct, or that I am missing something.
Has anybody encountered this issue before?
Is there a more detailed description on preprocessing longitudinal data in VBM?
How would I validate this result in absence of a healthy test-retest sample?
Thanks in advance,
Binod
|