The choice of how you do the coregistration should have little
systematic effect on how you spatially normalise (as long as the
images are all well aligned - use Check Reg to assess this).
How you actually spatially normalise the data will depend on what
works best. If there is good alignment between the functional and
structural scans (ie little distortion in the fMRI) then you can
estimate the warps from the structural images. If the alignment is
not so accurate, then it may be better to do it from the mean fMRI.
I prefer to use the Segment button for this spatial normalisation
(because it is more accurate). This will generate a parameter file
that encodes warps that can be applied to all scans of that subject
(providing all the rigid-registration stuff has worked well).
Best regards,
-John
On 10 November 2010 08:33, sarika cherodath <[log in to unmask]> wrote:
> Hi all,
> Is there any difference between the following methods of
> coregistration?
> (1) realigning all functional images--coregister: using realigned T1 image
> (rs) as "reference image", mean functional image (means) as "source image",
> all the other functional images (rs) for "other images"-- normalize:
> "rmeans" functional image and all other functional images (rrs) selected and
> normalized onto EPI.nii template
> (2) realign all functional--coregister: choose mean functional as ref.image
> and structural as source-- normalize: select coregistered structural and
> normalize onto EPI.nii template.
> Will the results obtained be the same in both cases??
> --
> Sarika Cherodath
> Graduate Student
> National Brain Research Centre
> Manesar, Gurgaon -122050
> India
>
>
|