On 8/27/2020 5:07 AM, Nathalie Mertens wrote:
External Email - Use Caution
Dear, Many thanks for these explanations! Further, I do have 4 additional questions linked to the mask option when applying RBV PVC in Freesurfer. (1)Could you please tell me _how_ the brainmask is generated when using the –auto-mask option, eg –-auto-mask 5 0.01. From which image is the generation of the mask started? (brain.mgz, brainmask.auto.mgz, brainmask.mgz, or something else?)
This not a brain mask but a whole head mask. It starts with gtmseg.mgz mapped into the PET space and binarized. It is then smoothed by 5 (your number) and the result is thresholded at 0.01 (your number). In the end, I have not found these numbers to be that important.
(2)I noticed that the amount of segments (used in RBV PVC) is reduced (compared to amount of roi’s in gtmseg) after using the mask. Can you explain this?
Some of the segments are very small making them hard to correction for partial volume effects; this can lead to very unrealistic numbers.
(3)Could you please tell me _when_ in the RBV PVC calculations this mask is applied?
Only at the very beginning to select the voxels that will participate in the PVC calculation.
(4)I guess this mask has nothing to do with the ‘masking rbv to brain’, which is applied in the end when running the RBV PVC? Can you elaborate on this final step?
--------- According to my notes: ---------------------------- // Reduce RBV to a tight bounding box around the brain by excluding anything // with "head" tissue type (hardcoded=5). This can greatly reduce // memory requirements.
Many thanks for your help. Kind regards, Nathalie Mertens On 8/19/2020 2:42 AM, Nathalie Mertens wrote: External Email - Use Caution Dear,
I am not sure if the questions below were sent correctly to you, sinceI did not receive any copy of it.
Please find my message below. Many thanks, looking forward to your reply. Kind regards, Nathalie Mertens sorry, did not see the original posting *Van: *Nathalie Mertens nathalie.mert...@kuleuven.be *Datum: *dinsdag 18 augustus 2020 om 09:33 *Aan: *"freesurfer@nmr.mgh.harvard.edu" freesurfer@nmr.mgh.harvard.edu *Onderwerp: *RBV PVC Dear, I do have some questions linked to RBV PVC implemented in Freesurfer:
- Could you please tell me how the RBV PVC is acquired? Is the RBV
PVC acquired based on Muller-Gartner (on a region-level, and based on only using gray- versus white matter), or when and how are the different regions from gtm_seg applied to obtain GTM?
See the RBV paper. Thomas, B.A., Erlandsson, K., Modat, M., Thurfjell,L., Vandenberghe, R., Ourselin, S.,Hutton, B.F., 2011. The importance of appropriate partial volumecorrection for PETquantification in Alzheimer's disease. Eur. J. Nucl. Med. Mol. Imaging38, 1104\u20131119.
- What would be the advantage of using your technique compared to
for instance sGTM? sGTM is ROI-only, RBV is voxelwise.
- Could you please tell me the difference in region-definition
between aparc+aseg and the gtm_seg? (I saw that the latter contains less regions compared to aparc+aseg.) What is the reason to use the latter compared to the aparc+aseg, based on the definitions of the regions? (I know that gtm_seg is used which has a high resolution, but not quite sure, apart from the resolution, why you are not using aparc+aseg for the RBV PVC.)
gtmseg.mgz is higher resolution (using the subvoxel surfaces). It alsohas pons and vermis defined. It has a few regions merged with others. It also has extracerebral structures.
- Is there a mask applied to the gtm_seg to exclude regions that are
located outside the brain? No, it is a full head model.
Many thanks for your answers. Kind regards, Nathalie _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer