External Email - Use Caution
Dear expert,
I am trying to coregister the Buckner cerebellar parcellation ( http://www.freesurfer.net/fswiki/CerebellumParcellation_Buckner2011) in all my sample. However the results that I got using mri_vol2vol are not so good (see attached). Indeed in most cases the parcellation is bigger than the anatomical cerebellum of my subjects. Do you have any tips?
Thanks
External Email - Use Caution
Hi Davide,
Can you elaborate more? You are trying to warp the Buckner parcellation from MNI parcellation to where?
What's the exact command that you used?
--Thomas
On Sun, Feb 9, 2020 at 2:03 AM Davide Momi momi.davide89@gmail.com wrote:
External Email - Use CautionDear expert,
I am trying to coregister the Buckner cerebellar parcellation ( http://www.freesurfer.net/fswiki/CerebellumParcellation_Buckner2011) in all my sample. However the results that I got using mri_vol2vol are not so good (see attached). Indeed in most cases the parcellation is bigger than the anatomical cerebellum of my subjects. Do you have any tips?
Thanks _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
External Email - Use Caution
Dear Thomas,
yes I am trying to warp the Buckner Parcellation to anatomical of all my subjects Please find below the command that I used:
mri_vol2vol --mov ../Buckner2011_7Networks_MNI152_FreeSurferConformed1mm_LooseMask.nii.gz \
--s 114823 \
--targ 114823/mri/brainmask.mgz \
--o Buckner2011_atlas_freesurfer_internal_space.nii.gz \
--interp nearest
Unfortunately the output is much bigger than the anatomical as I showed you in my previous email
Would you have any suggestions?
Thanks
Il giorno sab 8 feb 2020 alle ore 21:43 Thomas Yeo ythomas@csail.mit.edu ha scritto:
External Email - Use CautionHi Davide,
Can you elaborate more? You are trying to warp the Buckner parcellation from MNI parcellation to where?
What's the exact command that you used?
--Thomas
On Sun, Feb 9, 2020 at 2:03 AM Davide Momi momi.davide89@gmail.com wrote:
External Email - Use CautionDear expert,
I am trying to coregister the Buckner cerebellar parcellation ( http://www.freesurfer.net/fswiki/CerebellumParcellation_Buckner2011) in all my sample. However the results that I got using mri_vol2vol are not so good (see attached). Indeed in most cases the parcellation is bigger than the anatomical cerebellum of my subjects. Do you have any tips?
Thanks _______________________________________________ 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
External Email - Use Caution
Hi Davide,
Here's a previous email I sent to the list about warping Choi's striatum atlas to individual subject's striatum (which I have modified for the cerebellum). Let us know if this works.
1) Assuming you are quite happy with the freesurfer cerebellum parcellation in your subjects, then I am assuming freesurfer nonlinear registration (talairach.m3z) is working quite well. Talairach.m3z warps your subject to an internal freesurfer space (kinda like MNI305, but not quite). Let's say the freesurfer recon-all output is at <something>/SUBJECT_FS/
2) Run the MNI152 1mm template (the one from FSL) through recon-all. Recon-all will give you a Talairach.m3z that allows you to map the MNI152 1mm template to the internal freesurfer space. Let's say the freesurfer recon-all output is at <something>/MNI152_FS/
3) Then do the following:
a) Use mri_vol2vol to upsample the Buckner cerebellum atlas which is 2mm resolution to the 1mm MNI152 template:
mri_vol2vol --mov Buckner_atlas.nii.gz --targ MNI152/mri/norm.mgz
--regheader --o Buckner_atlas1mm.nii.gz --no-save-reg --interp nearest
Notice that I use norm.mgz of the MNI template rather than the original MNI template. norm.mgz is the 256 x 256 x 256 conformed version of the MNI template that recon-all puts through.
b) warp the Buckner_atlas1mm.nii.gz to freesurfer nonlinear volumetric space:
setenv SUBJECTS_DIR <something> mri_vol2vol_used --mov Buckner_atlas1mm.nii.gz --s MNI152_FS --targ
$FREESURFER_HOME/average/mni305.cor.mgz --m3z talairach.m3z --o Buckner_atlas_freesurfer_internal_space.nii.gz --interp nearest
c) warp the Buckner_atlas_freesurfer_internal_space.nii.gz to your subject:
setenv SUBJECTS_DIR <something> mri_vol2vol --mov $SUBJECTS_DIR/SUBJECT_FS/mri/norm.mgz --s SUBJECT_FS
--targ Buckner_atlas_freesurfer_internal_space.nii.gz --m3ztalairach.m3z --o Buckner_atlas_subject.nii.gz --interp nearest --inv-morph
This is not optimal because of the double interpolation. You might want to use the MNI template instead of the Buckner_atlas to test the above, so you can check the goodness of the warp. The final warped MNI template should hopefully look identical to your subject. If that works, then use the Buckner_atlas. Note that mri_vol2vol does not work properly for talairach.m3z below version 5, so you should use version 5x mri_vol2vol.
Regards, Thomas
On Sun, Feb 9, 2020 at 9:39 PM Davide Momi momi.davide89@gmail.com wrote:
External Email - Use CautionDear Thomas,
yes I am trying to warp the Buckner Parcellation to anatomical of all my subjects Please find below the command that I used:
mri_vol2vol --mov ../Buckner2011_7Networks_MNI152_FreeSurferConformed1mm_LooseMask.nii.gz \
--s 114823 \ --targ 114823/mri/brainmask.mgz \ --oBuckner2011_atlas_freesurfer_internal_space.nii.gz \
--interp nearestUnfortunately the output is much bigger than the anatomical as I showed you in my previous email
Would you have any suggestions?
Thanks
Il giorno sab 8 feb 2020 alle ore 21:43 Thomas Yeo ythomas@csail.mit.edu ha scritto:
External Email - Use CautionHi Davide,
Can you elaborate more? You are trying to warp the Buckner parcellation from MNI parcellation to where?
What's the exact command that you used?
--Thomas
On Sun, Feb 9, 2020 at 2:03 AM Davide Momi momi.davide89@gmail.com wrote:
External Email - Use CautionDear expert,
I am trying to coregister the Buckner cerebellar parcellation ( http://www.freesurfer.net/fswiki/CerebellumParcellation_Buckner2011) in all my sample. However the results that I got using mri_vol2vol are not so good (see attached). Indeed in most cases the parcellation is bigger than the anatomical cerebellum of my subjects. Do you have any tips?
Thanks _______________________________________________ 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
External Email - Use Caution
Thanks Thomas,
it worked perfectly
Best
Il giorno dom 9 feb 2020 alle ore 09:20 Thomas Yeo ythomas@csail.mit.edu ha scritto:
External Email - Use CautionHi Davide,
Here's a previous email I sent to the list about warping Choi's striatum atlas to individual subject's striatum (which I have modified for the cerebellum). Let us know if this works.
- Assuming you are quite happy with the freesurfer cerebellum
parcellation in your subjects, then I am assuming freesurfer nonlinear registration (talairach.m3z) is working quite well. Talairach.m3z warps your subject to an internal freesurfer space (kinda like MNI305, but not quite). Let's say the freesurfer recon-all output is at <something>/SUBJECT_FS/
- Run the MNI152 1mm template (the one from FSL) through recon-all.
Recon-all will give you a Talairach.m3z that allows you to map the MNI152 1mm template to the internal freesurfer space. Let's say the freesurfer recon-all output is at <something>/MNI152_FS/
- Then do the following:
a) Use mri_vol2vol to upsample the Buckner cerebellum atlas which is 2mm resolution to the 1mm MNI152 template:
mri_vol2vol --mov Buckner_atlas.nii.gz --targ MNI152/mri/norm.mgz
--regheader --o Buckner_atlas1mm.nii.gz --no-save-reg --interp nearest
Notice that I use norm.mgz of the MNI template rather than the original MNI template. norm.mgz is the 256 x 256 x 256 conformed version of the MNI template that recon-all puts through.
b) warp the Buckner_atlas1mm.nii.gz to freesurfer nonlinear volumetric space:
setenv SUBJECTS_DIR <something> mri_vol2vol_used --mov Buckner_atlas1mm.nii.gz --s MNI152_FS --targ
$FREESURFER_HOME/average/mni305.cor.mgz --m3z talairach.m3z --o Buckner_atlas_freesurfer_internal_space.nii.gz --interp nearest
c) warp the Buckner_atlas_freesurfer_internal_space.nii.gz to your subject:
setenv SUBJECTS_DIR <something> mri_vol2vol --mov $SUBJECTS_DIR/SUBJECT_FS/mri/norm.mgz --s SUBJECT_FS
--targ Buckner_atlas_freesurfer_internal_space.nii.gz --m3ztalairach.m3z --o Buckner_atlas_subject.nii.gz --interp nearest --inv-morph
This is not optimal because of the double interpolation. You might want to use the MNI template instead of the Buckner_atlas to test the above, so you can check the goodness of the warp. The final warped MNI template should hopefully look identical to your subject. If that works, then use the Buckner_atlas. Note that mri_vol2vol does not work properly for talairach.m3z below version 5, so you should use version 5x mri_vol2vol.
Regards, Thomas
On Sun, Feb 9, 2020 at 9:39 PM Davide Momi momi.davide89@gmail.com wrote:
External Email - Use CautionDear Thomas,
yes I am trying to warp the Buckner Parcellation to anatomical of all my subjects Please find below the command that I used:
mri_vol2vol --mov ../Buckner2011_7Networks_MNI152_FreeSurferConformed1mm_LooseMask.nii.gz \
--s 114823 \ --targ 114823/mri/brainmask.mgz \ --oBuckner2011_atlas_freesurfer_internal_space.nii.gz \
--interp nearestUnfortunately the output is much bigger than the anatomical as I showed you in my previous email
Would you have any suggestions?
Thanks
Il giorno sab 8 feb 2020 alle ore 21:43 Thomas Yeo ythomas@csail.mit.edu ha scritto:
External Email - Use CautionHi Davide,
Can you elaborate more? You are trying to warp the Buckner parcellation from MNI parcellation to where?
What's the exact command that you used?
--Thomas
On Sun, Feb 9, 2020 at 2:03 AM Davide Momi momi.davide89@gmail.com wrote:
External Email - Use CautionDear expert,
I am trying to coregister the Buckner cerebellar parcellation ( http://www.freesurfer.net/fswiki/CerebellumParcellation_Buckner2011) in all my sample. However the results that I got using mri_vol2vol are not so good (see attached). Indeed in most cases the parcellation is bigger than the anatomical cerebellum of my subjects. Do you have any tips?
Thanks _______________________________________________ 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
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Davide Momi -
Thomas Yeo