Dear Freesurfers,
I am using mri_label2label in Fs v5.0.0 to map a label between subjects. I have some problems in that the target label shows little holes while the location of the target label seems to correspond rather well with the source label. Checking these holes backward to the source subject using the function tools/send to subject in tksurfer ... brings the saved point in the source label.
I think this problem has been reported in the list before (see copy below) but unfortunately I do not find a solution that removes the holes. In my case I see this now and then both for ROIs that I drew myself in tksurfer but also pre-defined ROIs that were already there (e.g., rh.G_and_S_cingul-Mid-Ant.label). What might go wrong here? I am using --regmethod surface and trgsurf white.
I wondered further how one may check how well mri_label2label worked for each subject. Is there a way to test that source and target labels have the same location in terms of for example gyri and sulci? Are there cases that mri_label2label cannot work?
Thank you very much for your help, any suggestions appreciated!
Ruthger Righart (PhD)
Institute for Stroke and Dementia Research (ISD) Klinikum der Universität München Max-Lebsche-Platz 30 81377 Munich | Germany
Ruthger.Righart@med.uni-muenchen.de
http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg15125.html
Re: [Freesurfer] Mri_label2label problem? Douglas N Greve Tue, 21 Sep 2010 06:47:09 -0700 what does the original label look like? Can you convert one of the predefined labels (eg, lh.V1.label) to see if the same thing happens?
doug
Pye, Brandon (NIH/NIMH) [F] wrote:
Hey guys,
We are using mri_label2label to convert a label created from a cluster (using mri_surfcluster) from an average subject to each individual subject. We have 6 individual labels that we are converting. All of the analysis is done using surface data. I use the command in the generalized form listed below:
Mri_label2label -srcsubject fsaverage -srclabel cluster1.label -trg subject Subject -trglabel Subject_cluster1.label -regmethod surface -hemi lh
5 of the 6 labels transfer well, but a 6th label translated to the correct anatomical location, but the number of labeled vertices is greatly reduced after the translation. This problem persists across all subjects. There doesn't appear to be much interior to the label, just a lot of the outline of the label (i.e. If the outline of the label is yellow and the interior is blue, I see a lot of yellow and not much blue, as if there are a bunch of small "islands"). The region on them most inferior part of the lateral surface approximately where the temporal and occipital lobes meet.
Any ideas what might be going on?
Thanks,
Brandon Pye Post-Baccalaureate IRTA Laboratory of Brain and Cognition, NIMH/NIH/DHHS Building 10, Room 4C214 10 Center Dr MSC 1366 Bethesda, MD 20892 Phone: (301) 402-3893 Fax: (301) 402-0921 brandon....@nih.gov
Can you send me the label? doug
Righart, Ruthger Dr. wrote:
Dear Freesurfers,
I am using mri_label2label in Fs v5.0.0 to map a label between subjects. I have some problems in that the target label shows little holes while the location of the target label seems to correspond rather well with the source label. Checking these holes backward to the source subject using the function tools/send to subject in tksurfer ... brings the saved point in the source label.
I think this problem has been reported in the list before (see copy below) but unfortunately I do not find a solution that removes the holes. In my case I see this now and then both for ROIs that I drew myself in tksurfer but also pre-defined ROIs that were already there (e.g., rh.G_and_S_cingul-Mid-Ant.label). What might go wrong here? I am using --regmethod surface and trgsurf white.
I wondered further how one may check how well mri_label2label worked for each subject. Is there a way to test that source and target labels have the same location in terms of for example gyri and sulci? Are there cases that mri_label2label cannot work?
Thank you very much for your help, any suggestions appreciated!
Ruthger Righart (PhD)
Institute for Stroke and Dementia Research (ISD) Klinikum der Universität München Max-Lebsche-Platz 30 81377 Munich | Germany
Ruthger.Righart@med.uni-muenchen.de
http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg15125.html
Re: [Freesurfer] Mri_label2label problem? Douglas N Greve Tue, 21 Sep 2010 06:47:09 -0700 what does the original label look like? Can you convert one of the predefined labels (eg, lh.V1.label) to see if the same thing happens?
doug
Pye, Brandon (NIH/NIMH) [F] wrote:
Hey guys,
We are using mri_label2label to convert a label created from a cluster (using mri_surfcluster) from an average subject to each individual subject. We have 6 individual labels that we are converting. All of the analysis is done using surface data. I use the command in the generalized form listed below:
Mri_label2label -srcsubject fsaverage -srclabel cluster1.label -trg subject Subject -trglabel Subject_cluster1.label -regmethod surface -hemi lh
5 of the 6 labels transfer well, but a 6th label translated to the correct anatomical location, but the number of labeled vertices is greatly reduced after the translation. This problem persists across all subjects. There doesn't appear to be much interior to the label, just a lot of the outline of the label (i.e. If the outline of the label is yellow and the interior is blue, I see a lot of yellow and not much blue, as if there are a bunch of small "islands"). The region on them most inferior part of the lateral surface approximately where the temporal and occipital lobes meet.
Any ideas what might be going on?
Thanks,
Brandon Pye Post-Baccalaureate IRTA Laboratory of Brain and Cognition, NIMH/NIH/DHHS Building 10, Room 4C214 10 Center Dr MSC 1366 Bethesda, MD 20892 Phone: (301) 402-3893 Fax: (301) 402-0921 brandon....@nih.gov
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 -
Righart, Ruthger Dr.