I would suggest looking at the Klein et al paper in NeuroImage, which
should show you how to decide.
When you do any kind of analysis, the results you get will depend on
the models you used. The accuracy of your interpretation depends on
the accuracies of the models. The evaluations by Klein et al showed
that Dartel was more accurate (when compared against manually drawn
regions) than the older registration approaches in SPM. A more recent
evaluation of Dartel (using a better segmentation approach) in my
NeuroImage paper on Geodesic Shooting showed that Dartel could perform
better than it did in the Klein et al paper.
On 29 June 2011 10:18, mmkleung <[log in to unmask]> wrote:
> Dear all and Dr. Gaser,
> Regarding the VBM8 toolbox, I would like to ask how to decide which to use, the high-dimensional Dartel or the low-dimensional SPM normalization? Does it depend on the sample size or there are some other considerations.
> Thanks for your reply!
> Best wishes,