Re: [Freesurfer] custom ROI as a reference region for PET

Hi Lauri, yes, that's fine. It should give the same answer since you are not using PVC or accounting for TFE. doug

On 3/26/2025 11:14 AM, Lauri Tuominen wrote:

External Email - Use Caution

Dear PETsurfers,

I am trying to use PETsurfer and MRTM2 to analyze a SynVesT1 PET scan with centrum semiovale as a reference region. I have created a binary mask for centrum semiovale.

I guess my issue is that I don't understand what mri_gtmpvc does if I don't want to do partial volume correction.

So my question is if it's ok to extract the time series from the reference region using eg. mri_segstats and use it in the subsequent commands something like so:

mri_gtmpvc --i ${pet_dir}/mc_${pet_file} --reg 
${pet_dir}/p2mri.reg.lta --seg gtmseg.mgz \
--default-seg-merge --auto-mask 1.01 --no-tfe --o 
${pet_dir}/gtmpvc_hb.output \
--km-hb 11 12 13 50 51 52 --no-rescale --max-threads-minus-1

mri_segstats --i ${pet_dir}/mc_${pet_file} --reg 
${pet_dir}/p2mri.reg.lta --seg 
$SUBJECTS_DIR/test_subject/mri/cs_eroded.nii.gz --id 1 --avgwf 
${pet_dir}/ref.km.txt

mri_glmfit --y ${pet_dir}/gtmpvc_hb.output/km.hb.tac.nii.gz --mrtm1 
${pet_dir}/ref.km.txt time.dat --o mrtm1 --no-est-fwhm --nii.gz

Thank you so much!

Lauri Tuominen

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer