Could someone please explain to me the difference between mri_binarize and mri_extract_label?
I had been using mri_extract_label to create ROI masks of cerebral white matter (2 and 41). I recently became aware that this would give me a label with 0 and 128. I tried mri_binarize (with the same matchvals 2 and 41) and it gave me a label of 0 and 1. However, when comparing the two in tkmedit, it appears as though the mask I created using mri_binarize includes more of the corpus callosum region in mid-sagittal slices. I've tried this on a couple different subjects, so I do not think it is limited to one person. Is this a typical difference between the two commands? And if so, which one is recommended for extracting white matter masks?
Also to note that before running either of these commands, I also convert the aseg.mgz to the subject's native space (using rawavg.mgz as a template). The commands I ran are the following:
mri_label2vol --seg aseg.mgz --temp rawavg.mgz --o aseg-in-rawavg.mgz --regheader aseg.mgz
mri_extract_label aseg-in-rawavg.mgz 2 41 cerebral_wm.nii
mri_binarize --i aseg-in-rawavg.mgz --match 2 41 --wm --o cerebral_wm_binarize.nii
On another note, I was wondering if there is a way to create a probability map for the white matter. That is, instead of getting a binary (0s and 1s) mask, I would get the probability of each voxel being part of the white matter ROI with values ranging from 0 to 1.
Thanks for your help,
Vanessa Palzes Staff Research Associate Brain Imaging & EEG Lab San Francisco VA Medical Center 4150 Clement Street San Francisco, CA 94121 Tel: 415-221-4810 ext. 6155
When you add --wm, you are telling mri_binarize is give you all of white matter (this include CC and cerebellar WM and hypointensities). When you pass just 2 and 41, you get only cerebral WM. doug
Palzes, Vanessa wrote:
Could someone please explain to me the difference between mri_binarize and mri_extract_label?
I had been using mri_extract_label to create ROI masks of cerebral white matter (2 and 41). I recently became aware that this would give me a label with 0 and 128. I tried mri_binarize (with the same matchvals 2 and 41) and it gave me a label of 0 and 1. However, when comparing the two in tkmedit, it appears as though the mask I created using mri_binarize includes more of the corpus callosum region in mid-sagittal slices. I've tried this on a couple different subjects, so I do not think it is limited to one person. Is this a typical difference between the two commands? And if so, which one is recommended for extracting white matter masks?
Also to note that before running either of these commands, I also convert the aseg.mgz to the subject's native space (using rawavg.mgz as a template). The commands I ran are the following:
mri_label2vol --seg aseg.mgz --temp rawavg.mgz --o aseg-in-rawavg.mgz --regheader aseg.mgz
mri_extract_label aseg-in-rawavg.mgz 2 41 cerebral_wm.nii
mri_binarize --i aseg-in-rawavg.mgz --match 2 41 --wm --o cerebral_wm_binarize.nii
On another note, I was wondering if there is a way to create a probability map for the white matter. That is, instead of getting a binary (0s and 1s) mask, I would get the probability of each voxel being part of the white matter ROI with values ranging from 0 to 1.
Thanks for your help,
Vanessa Palzes Staff Research Associate Brain Imaging & EEG Lab San Francisco VA Medical Center 4150 Clement Street San Francisco, CA 94121 Tel: 415-221-4810 ext. 6155
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Douglas N Greve -
Palzes, Vanessa