Hi Nadine,
> We are trying to normalize ROIs (created using BIT/Marsbar) to a
> customized child template in SPM5. As the source image as well as the
> image to write we choose our ROI, for the Template image we select our
> customized child template (created with TOM).
I think the problem may be that, one, your ROI is already in normalized
space, and if you try to normalize it (again), things go astray. Two, a
(likely small) region of interest does not have much similarity with the
whole-brain TOM template, so matching them is difficult in either case.
What we have done in the past (Neuroimage 2003, 20: 202-215) is to match
the brain the ROIs are derived of (the AAL brain in our case) to our
custom template, using a high-dimensional warping without an affine
normalization (which, in spm, assumes the target brain to be quite a bit
bigger). We used the deformation toolbox and then applied those
parameters to the ROI. We also smoothed and then binarized the masks a
bit in order to account for possibly inexact normalization, but that
would depend on the algorithm you use.
Hope this helps,
Marko
--
=====================================================================
Marko Wilke (Dr.med./M.D.)
[log in to unmask]
Universitäts-Kinderklinik University Children's Hospital
Abt. III (Neuropädiatrie) Dept. III (Pediatric neurology)
Hoppe-Seyler-Str. 1, D - 72076 Tübingen
Tel.: (+49) 07071 29-83416 Fax: (+49) 07071 29-5473
=====================================================================
|