The templates are only incidental to the "Run DARTEL (create
Templates)" option. The important output is the parametrisation of
the relative shapes of the brains, which are encoded by the u_*.nii
files. You would normally use these to transform your data to the
same common space. See the manual chapter on Using Dartel for more
information.
Note that doing an SPM analysis of FA is not morphometry (as defined
http://medical-dictionary.thefreedictionary.com/morphometry and
http://medical-dictionary.thefreedictionary.com/form), so should not
really be referred to as VBM.
Also note that doing a simple SPM analysis to compare groups of warped
FA maps may not give the most easily interpretable findings. Many of
the statistically significant differences you find (depending on
whether you have enough subjects in the study) will simply be because
of the different brain shapes - rather than because of any differences
between the FA in homologous regions. There are ways of trying to
partition the information into different tissue types, so I'd suggest
looking at eg:
Jee Eun Lee, Moo K. Chung, Mariana Lazar, Molly B. DuBray, Jinsuh Kim,
Erin D. Bigler, Janet E. Lainhart, Andrew L. Alexander. A study of
diffusion tensor imaging by tissue-specific, smoothing-compensated
voxel-based analysis. NeuroImage, Volume 44, Issue 3, 1 February 2009,
Pages 870-883
Alternatively, there is always TBSS.
Best regards,
-John
On 18 July 2012 09:14, Alejandro Cosa Liñán <[log in to unmask]> wrote:
> Dear all,
>
> I am using SPM for VBM-Style FA.
>
> 21 dwMRI and its high-resolution structural MRI was acquired. After scaling dimension (x10), creating FA maps corregistering dwMRI and structural MRI, normalisation with DARTEL is performing.
>
> Segmentation of different tissues is performed with "New Segment" with the option "Native + DARTEL Imported". After this, I run "Run DARTEL (create Templates)" and 21 "rc1*" and 21 "rc2*" images are selected for create templates. Templates are succesfsfully created.
>
> After, templates are created, how should I to continue? How should I apply calculated transformation to my FA maps? I think about reslicing FA maps to "rc*" space, but it is not work.
>
> Thanks,
>
> Alejandro Cosa Liñán
> PhD sudent.
> Centro de Biomateriales e Ingeniería Tisular,
> Universidad Politécnica de Valencia, Spain
|