As far as I'm aware, there shouldn't be any real issues unless you
also spatially smooth the native-space data. Any differences you see
between doing it the more conventional way and this way will be the
result of a slight change in the estimation of the variance
components. Otherwise, whether you warp a linear combination of
images, or generate a linear combination of warped images, should give
the same results.
Variance components estimated from spatially smoothed warped images
will be a little bit different from those estimated from the original
images, which will have an effect on how the fMRI scans are linearly
combined. I don't know how small these differences are likely to be.
Best regards,
-John
On 2 March 2013 15:17, Jeff Browndyke <[log in to unmask]> wrote:
> SPMers,
>
> Any inherent problems or issues associated with performing 1st level analyses in native space and then warping the resulting .con images with either standard normalization or DARTEL method?
>
> On a related note, would this type of processing scenario cause a problem with GIFT or MANCOVAN?
>
> I have longitudinal data (baseline, 6wk, 1yr) and thought it might be nice to retain the subjects' data in native space for time x variable analyses on an individual level for visualization, but then take their .con images into MNI space for group x time x variable comparisons.
>
> Thanks,
> Jeff
>
> -----------------------------------------------------------------------
> Jeff Browndyke, Ph.D.
> Clinical & Research Neuropsychologist
> Durham VA & Duke University Medical Centers
>
> [log in to unmask] / [log in to unmask]
> office: (919) 286-0411 ext. 4656
> cell: (336) 264-4222
|