There are many things that I would like to eventually include within SPM, but
because there is only one of me, then I can only suggest workarounds at the
moment. Ideally, processing all the images of a single subject would be done
together within the same model, but this is not currently possible. I don't
know what the best procedure is likely to be, but you could try:
1) Coregister your volumes (within subject). Reslicing is not necessary.
2) Segment them, producing c*.img files.
3) Take the *_seg_sn.mat file of one of the subjects, and use this to create
spatially normalised (and modulated) versions of all the images of each
subject (with the Normalise button).
There will be differences between how the tissue probability maps are
overlayed on the scans (because segmentation estimates different warps for
each scan), and I don't know how this influences the outcome.
Best regards,
-John
> I am planing a logitudinal VBM analysis with SPM5.
> I have read all emails in the archives concerning longitudinal VBM
> but there remains some questions:
>
> I intend to produce customised T1 template and corresponding GM/WM/CSF a
> priori maps
> (because the brains are from AD patients). I know that I have to
> coregister(without resclicing)
> the late images on the early images. But where should I incorporate the
> coregistration?
>
> 1. Should I coregister the native T1 images before I start to create the
> customised template and a priori maps?
> 2. Should I start with the normalisation and segmentation to create the T1
> template and a priori maps and then coregister the late images on the early
> images before I start to normalise and segment the native images again but
> with the customised template and a priori maps?
> 3. I have four different time points to coregister. The reference image is
> the baseline image (earliest one).
> Is it OK when I select the latest image as source image and the remaining
> two images as "Other images" or must I coregister all three later images to
> the earliest image separately? I think that the results from both
> procedures will be the same.
|