actually, if you convert the aparc annotations to a bunch of labels, you can run:
mri_cnr -label <left hemi label name> <right hemi label name> ...
and it will limit the calculations to that label. FOr example:
mri_cnr -label ~/links/subjects/bruce.dev//label/lh.cortex.label ~/links/subjects/bruce.dev//label/rh.cortex.label ~/links/subjects/bruce.dev/surf ~/links/subjects/bruce.dev/mri/norm.mgz reading lh and rh labels from /homes/4/fischl/links/subjects/bruce.dev//label/lh.cortex.label and /homes/4/fischl/links/subjects/bruce.dev//label/rh.cortex.label processing MRI volume /homes/4/fischl/links/subjects/bruce.dev/mri/norm.mgz... white = 91.0+-9.7, gray = 57.8+-20.2, csf = 33.6+-15.4 gray/white CNR = 2.204, gray/csf CNR = 0.910 lh CNR = 1.557 white = 90.4+-10.0, gray = 57.6+-19.7, csf = 34.6+-15.2 gray/white CNR = 2.196, gray/csf CNR = 0.851 rh CNR = 1.523 total CNR = 1.540
cheers Bruce
On Wed, 3 Apr 2019, neuroimage analyst wrote:
External Email - Use Caution
Thanks, Dr. Greve. Where can I find the correspondence between the vertices in ?h.w-g.pct.mgh and the parcellations?
On Wed, Apr 3, 2019 at 2:41 PM Greve, Douglas N.,Ph.D. DGREVE@mgh.harvard.edu wrote: you can get the contrast from surf/?h.w-g.pct.mgh, which you can average over each parcellation. Not sure where you would get the noise term from.
On 4/3/19 5:23 PM, neuroimage analyst wrote: > > External Email - Use Caution > > I apologize, Dr. Fischl, that I wasnt clear. > > We are using Desikan-Killany atlas that has 68 surfaces defined on the > atlas, 34 on each hemisphere. > > I might be wrong, but I was thinking that there should be a way to > define CNR of orbitofrontal cortex (OFC), for example. I know that we > can get CNR of the whole brain gray/white/CSF but is it possible to > get the gray/white/CSF CNR of OFC only? If no, then we report the > difference in cortical thickness as we observe. If yes, we want to > regress or potentially discard the worst CNR subjects and recompute > the statistics. the hope is we can get 68 CNRs for each surface in the > Desikan-Killany atlas.--Kindly let me know your suggestions and thoughts. > > Thanks > > On Wed, Apr 3, 2019 at 2:11 PM Bruce Fischl > <fischl@nmr.mgh.harvard.edu <mailto:fischl@nmr.mgh.harvard.edu>> wrote: > > can you clarify what you mean? What 68 surfaces do you mean? And > what do > you mean by CNR of a surface? > On Wed, 3 Apr 2019, neuroimage analyst wrote: > > > > > External Email - Use Caution > > > > Hi, > > We were interested in measuring CNR of each of the 68 surfaces > in Desikan-Killany atlas and compare > > between the groups to verify if our results are being biased by > CNRs in the group? > > > > I came across mri_cnr and mris_ms_surface_cnr binaries in the > FreeSurfer tools and were wondering > > how to loop it through each subject in the subject_list.txt file > (each column in subject_list.txt is > > the subject id) so that we get a text file (similar to ?cortical > thickness txt file generated using > > aparcstats2table) that has # of subjects * # of regions txt file > and the entries within each matrix > > cell is the CNR value corresponding to surface X of Subject Y. > > > > We will greatly appreciate any help or pointers regarding this. > > > > Thanks > > > > Regards > > > > --VM > > > >_______________________________________________ > Freesurfer mailing list > Freesurfer@nmr.mgh.harvard.edu <mailto: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