Dear SPM community,
The issue of flipping images for voxel-based morphometry has been discussed on this list. Still I am wondering what might be the ideal way to study a lateralized disorder (e.g. temporal lobe epilepsy with epileptic focus on the right or left side) with DARTEL-VBM.
I want to combine left- and right sided patients (N=25) and compare them to another group of patients (N=23), and healthy controls (N=29).
What do you think of the following workflow?
1. Creation of symmetrical tissue probability maps (TPMs) for grey, white & csf (using SPM default files):
a. Symm_grey = (grey+flipped_grey)/2
b. Symm_white …
c. Symm_csf …
2. Creation of symmetrical DARTEL template, using original and left-right flipped images of all subjects (patients + controls)
3. DARTEL normalization of grey matter (original and flipped files where needed)
Many thanks for any comments or literature hints
Franz
|