Dear Mike,
That would depend on your regions. The SPM PET template is based on H2[15O] PET - which should in principle yield blood flow data reasonably close to e.g. metabolism data or, in the case of dynamic receptor ligand PET, early frames. However, it isn't particularly great in terms of actually matching MNI/ICBM(152) space.
You can check for yourself using CHECKREG and the PET.nii versus T1.nii templates. I just had a quick look to refresh my memory and found large differences in the occipital poles, inferior temporal lobes, along the insula, and particularly around the orbitofrontal cortex.
Should that be the problem, you might be able to do something about it by creating your own templates. This should be more important the more your ligand and / or PET scanner characteristics are away from the data used for SPM's PET.nii (click the green HELP button and then TEMPLATES for further info).
Obviously if you describe your blobs in anatomical terms and relative to the template used, there isn't a problem either given your description of a good fit to the template - it's just that the PET.nii coordinates will be somewhat different from other "MNI/ICBM" coordinates.
Hope this helps,
All the best,
Alexander
On 27 Jan 2011, at 15:32, Mike Sugarman wrote:
> Hello SPMers,
>
> I'm working with PET data and normalized all of my scans to the SPM-provided template ("PET.nii"). However, when I run a basic model and fit it to the MIP, most of my regions of activation lie outside of the "glass brain". My activation matches properly to the template. I feel the template provided by SPM may be too large, does anybody else have similar issues or know of a fix?
>
> Thanks,
> -Mike Sugarman
> Wayne State University
|