Hello experts,
I've recently run into a case the gtmstats uptake (reference region: cerebellum) for the pericarlcarine ROI's are as low as 0.40, however computing the voxel wise mean using input.rescaled.nii.gz and mgx.ctxgm.nii.gx and aux/seg.nii.gz, the values are closer to 1, as expected:
| vol | psf | roi | gtmstats_uptake | mean_uptake | | --------------------- | --- | -------------------- | --------------- | ----------- | | input.rescaled.nii.gz | 0 | ctx-rh-pericalcarine | 0.807 | 1.226… | | input.rescaled.nii.gz | 0 | ctx-lh-pericalcarine | 0.758 | 1.083… | | input.rescaled.nii.gz | 6 | ctx-rh-pericalcarine | 0.409 | 1.197… | | input.rescaled.nii.gz | 6 | ctx-lh-pericalcarine | 0.470 | 1.058… | | mgx.ctxgm.nii.gz | 0 | ctx-rh-pericalcarine | 0.807 | 1.007… | | mgx.ctxgm.nii.gz | 0 | ctx-lh-pericalcarine | 0.758 | 0.888… | | mgx.ctxgm.nii.gz | 6 | ctx-rh-pericalcarine | 0.409 | 0.915… | | mgx.ctxgm.nii.gz | 6 | ctx-lh-pericalcarine | 0.470 | 0.809… |
This is using Freesurfer 6.0, and the PET to anatomical registration looks fine.
Any ideas why we get such bad values from the GTM method?
Thanks for any help, Ryan