> Dear SPM expert and John,
> I try to use my SPECT image and overlay it on the T1 MRI image under
> standardized coordination. First, I use my SPECT image and normalize to
> the template SPECT, which result in a 79*95*69 size SPECT image. Than I us
> thte single subject T1 MRI and normalize to standardized T1 MR. I use this
> SPECT image and overlay it on this T1 MRI template (single subject) and
> find the cortical area intensity in SPECT image did not over the cortical
> area in T1 MRI. In some area, such as parietal- temporal area, the
> cortical area in SPECT is about 5 mm smaller than T1 MR image. Why the
> SPECT image did not match to the T1 MR ? If I change to use the
> coregistration, it seem improve the result. but it is not the standardized
> coordination. so, how to do it?
If you have T1 weighted MRI and SPECT scans of the same subject, then these
should be registered together using the Coregister button. Rigid-body
registration only needs 6 parameters to be estimated, so is usually more
robust than doing spatial normalisation, which has many more parameters.
The templates released with SPM may also not be ideal for your images, as
(for example) the contrast in the T1 template may differ slightly from the
contrast in your own T1-weighted scans.
Therefore, a better approach would be to coregister your images together.
After this coregistration, you can estimate spatial normalisation parameters
from one of your images, and then apply these to the others. For example,
spatially normalising an image called my_spect.img to the SPECT template
will produce a set of warps saved in my_spect_sn.mat. The contents of this
file can then be applied to your T1 weighted image in order to warp it
in the same way.
Providing the coregistration is accurate, then the SPECT and T1 images should
be accurately matched together in normalised space.
> Besides, in Apr 10, Marenco Stefano write the bounding box in spm2b have
> some problem (template 91*109*91 and normalize to 92*110*92). Prof John
> wrote " manually specify the bounding box as -90:90, -126:90, -72:108),
> can you explain more detail about how to change it?
You can use the Defaults button, then select spatial normalisation, and
then defaults for writing normalised images. This will then go through
the various defaults required. When you get to selecting a bounding box,
you would specify "Customise", and then enter:
- 90 90
-126 90
-72 108
Best regards,
-John
--
Dr John Ashburner.
Functional Imaging Lab., 12 Queen Square, London WC1N 3BG, UK.
tel: +44 (0)20 78337491 or +44 (0)20 78373611 x4381
fax: +44 (0)20 78131420 http://www.fil.ion.ucl.ac.uk/~john
|