Sophie,
If these are indeed to be treated as different experiments (i.e.,
they will be run in completely separate statistical models), then
there is no reason why the three runs need to be realigned together.
Realigning them all at once would get them all into the same space,
but it would likely require more transformation than if you do it
separately (the SPM realignment for multiple sessions is a two step
process). IMHO it is better not to do the additional transformations
if they are truly separate.
As far as your anatomy is concerned, you should duplicate the image
and coregister it to the functional images for each experiment
separately for use in other processing steps.
That's my two cents. Perhaps the SPM gurus will have a different
opinion on the matter.
Amy
>Very quick and hopefully easy question. I have to reanalyze some data:
>experiments has 3 runs (a,b,c). when I'm realigning, do I need to do them
>together or separately? Each run is a different experiment and they do not
>need to be subtracted from each other. In preliminary analysis, there has
>been approximately 1-2mm total shifting from the beginning of a to the end
>of c.
>
>So: realign together (lots of time) or realign separately (faster)?
>
>note: only 1 3-dimensional high res picture is taken, used for a & b & c.
>
>does this make a difference?
>
>thanks ;-)
>
>sophie
|