Dear Richard,
the DARTEL approach can be applied using either an existing template (such as in VBM8) or by creating a customized template. Both of these options are available in the DARTEL toolbox. However for the latter approach you need a large number of representative brains to create your own template. A template should represent a population or sample and therefore a minimum number of images is required (please don't ask me how many ;-). The smaller the number of images is (that were used to create the template) the less is the advantage of creating this customized template.
Thus, my idea by providing a template with VBM8 was to simplify the process for most studies where not enough images are available. Another advantage of this existing template is that it is already in MNI space (no additional MNI transformation is necessary) and results are comparable between study if this template is used in other studies. The customized template is in its own (customized) DARTEL space that might also differ between studies. However, the advantage is that less deformations are necessary and the anatomy as well is intensity profiles are closer to your data.
Anyway, you can also create your own template using VBM8 and the DARTEL toolbox. Please follow the instructions at page 25 in the VBM8 manual.
Best,
Christian
PS: If I remember correctly John Ashburner mentioned that 50-100 brains are the minimum number to create your own template.
__________________________________
Christian Gaser, Ph.D.
Departments of Psychiatry and Neurology
Friedrich-Schiller-University of Jena
Jahnstrasse 3, D-07743 Jena, Germany
Tel: ++49-3641-934752 Fax: ++49-3641-934755
e-mail: [log in to unmask]
http://dbm.neuro.uni-jena.de
On Mon, 14 May 2012 16:14:41 -0700, Richard Binney <[log in to unmask]> wrote:
>Dear Christian,
>
>Can you clear something up for me? My questions are a little naive due to
>confusion......
>
>I'm trying to understand what aspect(s) of DARTEL are incorporated into
>VBM8...if a pre-existing DARTEL-made template is provided (in MNI space),
>how are you using the DARTEL algorithm? My understanding of DARTEL is that
>its core functionality is in creating the group-specfic template with
>equidistant (as far as possible) transforms from each subject to the
>template but if the template is already provided then all there is to do is
>calculate the warps to this template...is this done using a
>high-dimensional warp that is specific to DARTEL (is this Run DARTEL
>(existing templates?)? Could you not simply use the high dimensional warp
>utility with your DARTEL-made template? Would this be better? Does using a
>pre-exisiting template adequately fulfill the minimal distance objective of
>DARTEL? Is a template made of healthy controls suitable for highly
>atrophied brains such as those of dementia populations? Another core
>function of DARTEL is the simultaneous warping and smoothing - can this not
>be incorporated? I guess it would need a template not in MNI space and then
>use of the DARTEL norm to MNI function.
>
>Richard
>
|