Hello,
it has sometimes been proposed to use a WM template for a WM analysis,
using a similar pipeline as for GM.
Most people present only GM data in their VBM studies, so the problem does
not appear.
It is my impression that for WM the situation has been about 50/50 before
spm5, however, to me it
makes sense to use a WM template as the first goal is good intergroup
correspondance of
your variable of interest, and this is voxelwise WM.
In fact, the fine coarse subcortical WM structures follow the gyri quite
well, so
normalisation quality is often quite good as far as one can judge from
looking at the images.
It may make a difference, also, as the modulation step (I guess you are
using modulated data) and not
only the normalisation matrix are different of course. So is seems it is
worth trying it.
BUT:
In spm5 with the unified normalisation algorithm that is intrinsically
interwoven with segmentation,
there would be no difference as from all three priors only ONE
normalisation matrix is gained. So by using
spm5, you could have circumvented the whole issue.
best regards,
Philipp
Max Planck Institute of Psychiatry
NMR Research Group
Kraepelinsr. 2-10
80804 Munich
Mail: [log in to unmask]
Phone: 0049-89-30622-413
|