Hello Jeff,
it seems you first need the realignment step which produces
a mat file for evrey image of the time series. I addition,
a coregistration to your T1 is needed (either by the 'coreg'
function or manually).
Technically: If you use the coreg function it seems you could reslice the fMRI images to
the reference space of the T1 (reslice only function), and by this step 'freeze' the two prior registration steps
and get fMRI at the resolution of the original T1.
However, for a native fMRI analysis you could (and maybe should) stay in the fMRI
resolution or something neear to it like 2x2x2 -
the anatomical resolution (usuallzy round 1x1x1) may be a bit too high. As long, as the T1 and fMRI
are coregistered you can still use the T1 as background image for display purposes.
As I am not so much in fMRI: try to find from some other hints/opinions if reslicing to higher
resolutions could decrease your statistical power in some way.
br,
Philipp
|