> We have a basic question about the co-registration in SPM2, which has
> puzzled us for a while.
>
> We tried to coregister PET1 and PET2 (both were acquired sequentially, so
> should be considered aligned already) to the subject's MR image. Because
> PET2 has poor contrast, so we decide to use PET1 (source) to coregister to
> the MR (target) to get the transformation file (PET1.mat).
>
> Then, we use "coregister" funtion again to try to "reslice only" the PET1
> (source) and PET2 (others) images. What puzzled us is that since there
> doesn't exist a PET2.mat file, how SPM is going to reslice the PET2 image?
> Does SPM use the one from PET1.mat as the transformation parameter for BOTH
> PET1 and PET2?
The "other images" should be used to specify any image that is in register
with the "source image". The idea is that a transformation is estimated for
the source image and saved in its .mat file, and that same transformation is
also saved for the "other" images.
>
> Alternatively, coud we simply copy and rename PET1.mat to PET2.mat, and
> then reslice the two images? Would the results of rPET2.img be the same as
> the way in the previous paragraph?
In this case, you can just copy the .mat file from PET1.mat to PET2.mat.
Best regards,
-John
|