Hi,
I haven't done these things myself but I think that most people used raw
FA maps instead of the skeleton for VBM-like analyses outside FSL
(although a comparison of skeletonised and raw FA maps would be
interesting).
The original TBSS paper shows a comparison in figure 15:
http://www.ncbi.nlm.nih.gov/pubmed/16624579
These papers may also help:
http://www.ncbi.nlm.nih.gov/pubmed/12377137
http://www.ncbi.nlm.nih.gov/pubmed/15907311
Best,
Andreas
Antoinette Nicolle schrieb:
> Dear experts,
>
> I'd like to find out how well it could work to take my skeletonised FA images from TBSS in FSL, through to a VBM-style multiple regression analysis in SPM. Perhaps it would be best to use FSL for my stats, but please indulge me for a moment since I'm attempting to compare against some results I already have from a VBM analysis in SPM.
>
> Firstly I believe these FA skeletons will be in MNI space, since I aligned them to the FMRIB58_FA template during TBSS. At least they appear to be in the correct space when I use check reg in spm. Could I be missing anything important here?
>
> Secondly, due to the high resolution (1x1x1mm)and lack of smoothing of these images, is it appropriate to smooth the skeletons (perhaps by just a small kernel, e.g. 4mm) prior to performing the SPM group analysis?
>
> I appreciate your advice, especially concerning any problems with this that I may not have anticipated.
>
> Many thanks,
> Antoinette
>
>
>
--
Andreas Hahn, MSc PhD
Functional, Molecular & Translational Neuroimaging
Department of Psychiatry and Psychotherapy
Medical University of Vienna, Austria
Phone: +43-1-40400-2320
Email: [log in to unmask]
Web: http://www.meduniwien.ac.at/neuroimaging/
|