Hello Everyone,
We are trying to compare structural differences between two groups using
VBM under SPM5. Group 1 had 16 subjects and Group 2 had 29 subjects. The
high resolution T1 images were Segmented and normalized with a modulate
option using the unified segmentation function in SPM5. We then smoothed
them with the FWHM=8*8*8. The preprocessed images seemed ok. Two sample
test was used to compare differences between the groups. However, the
results didn’t make any sense. Especially, there were many clusters of
voxels located outside of the brain. There was no any error message on
Matlab window during the process. An example of the preprocessed T1 images
and the statistic map are attached. Can anyone please take a look at the
images and the statistic map and help us to get an idea to fix the
problem? Any comments and suggestions will be greatly appreciated.
Thank you,
Jiang Xu
NIH/NIDCD
|