That's a good question, but I don't have a good answer. Registration
accuracy does seem to be improved (over older SPM registration
approaches) by using DARTEL (Klein et al, NeuroImage). I hope that
increased accuracy will lead to homologous functional regions lining up
a bit better, which should improve your t statistics.
Whether or not there is improvement in p values for PET images will
depend on a number of things, one of them being how blurred the PET data
are. If you have broad point-spread functions, then you may not see so
much improvement. You'll also need to ensure accurate alignment between
the PET and MRI data.
Best regards,
-John
On Tue, 2009-10-20 at 16:10 +0000, Gabriel Go.Es. wrote:
>
> hello SPMers
>
> i have some doubts for you about SPM
>
> i have FDG-PET images PVE corrected, and i want to do my statistical
> analysis on SPM, i have worked with VBM early on SPM, my questions
> are: should i use the DARTEL tools to work my PET images? is this tool
> going to make better in any way my statistical approaches? or is not
> nessesary to use it at all?
>
> any help will be appreciated
>
> Best Regards
> Gabriel
>
>
>
>
>
>
>
> ______________________________________________________________________
> Todo el espacio y cuidado que merecen tus fotos digitales lo tienes en
> Windows Live Fotos. ¡Pruébalo!
--
John Ashburner <[log in to unmask]>
|