Hippocampal/Amygdala subfield segmentation error - Anterior Amygdala Area missing in FreeView
External Email - Use Caution
Hi FreeSurfer Experts,
I'm attempting to QC the hippocampal/amygdala segmentation in FreeView. When looking at ?h.hippoAmygLabels-T1.v21.FS60.mgz I can see that a few subjects are either missing a mask of the anterior amygdaloid area (AAA) for a single hemisphere or the AAA is extremely small. However, in both circumstances the volumetric value for the AAA in the output (?h.amygNucVolumes-T1.v21.txt) looks normal. In the cases where no AAA mask is visible, if you check the box "Show existing labels only", the label "7010 Anterior-amygdaloid-area" does not even appear for that hemisphere.
AAA is the only structure where this is occurring. There are other structures (such as the medial nucleus) that appear small in FreeView, but still have a normal volumetric value. However, no other structure is completely missing.
Note: In all of these cases aseg looks fine. And T1s were processed using freesurfer-Darwin-OSX-stable-pub-v6.0.0-2beb96c, but we used freesurfer-darwin-macOS-7.1.0-20200511-813297b to run with hippocampal/amygdala segmentation (segmentHA_T1.sh).
Thank you for you time and help,
Sarah Skinner Research Specialist University of Wisconsin - Madison
Thanks for the interest in our tool, Sarah! With the smallest nuclei, it is possible that nucleus X has non-zero probability for a bunch of voxels, but does not win any of those voxels (i.e., there’s another label with larger probability). If that happens, the volume will be non-zero (as you have some volume coming from the non-zero probabilities) but no single label will be labeled as X. Does this make sense? Cheers, /Eugenio
-- Juan Eugenio Iglesias http://www.jeiglesias.com
From: freesurfer-bounces@nmr.mgh.harvard.edu on behalf of Sarah Skinner seskinner@wisc.edu Reply-To: Freesurfer support list Freesurfer@nmr.mgh.harvard.edu Date: Friday, January 13, 2023 at 11:46 AM To: "Freesurfer@nmr.mgh.harvard.edu" Freesurfer@nmr.mgh.harvard.edu Subject: [Freesurfer] Hippocampal/Amygdala subfield segmentation error - Anterior Amygdala Area missing in FreeView
External Email - Use Caution Hi FreeSurfer Experts,
I'm attempting to QC the hippocampal/amygdala segmentation in FreeView. When looking at ?h.hippoAmygLabels-T1.v21.FS60.mgz I can see that a few subjects are either missing a mask of the anterior amygdaloid area (AAA) for a single hemisphere or the AAA is extremely small. However, in both circumstances the volumetric value for the AAA in the output (?h.amygNucVolumes-T1.v21.txt) looks normal. In the cases where no AAA mask is visible, if you check the box "Show existing labels only", the label "7010 Anterior-amygdaloid-area" does not even appear for that hemisphere.
AAA is the only structure where this is occurring. There are other structures (such as the medial nucleus) that appear small in FreeView, but still have a normal volumetric value. However, no other structure is completely missing.
Note: In all of these cases aseg looks fine. And T1s were processed using freesurfer-Darwin-OSX-stable-pub-v6.0.0-2beb96c, but we used freesurfer-darwin-macOS-7.1.0-20200511-813297b to run with hippocampal/amygdala segmentation (segmentHA_T1.sh).
Thank you for you time and help,
Sarah Skinner Research Specialist University of Wisconsin - Madison
External Email - Use Caution
I think I understand what you are saying, but I'm uncertain how to best use this information. Do you have any guidance on how we should use this in analysis and publication? How comfortable should we be sharing this data publicly? What caveats should we include when sharing?
Sarah Skinner Research Specialist University of Wisconsin - Madison ________________________________ From: freesurfer-bounces@nmr.mgh.harvard.edu freesurfer-bounces@nmr.mgh.harvard.edu on behalf of Iglesias Gonzalez, Juan E. JIGLESIASGONZALEZ@mgh.harvard.edu Sent: Friday, January 13, 2023 12:46 PM To: Freesurfer support list Freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Hippocampal/Amygdala subfield segmentation error - Anterior Amygdala Area missing in FreeView
Thanks for the interest in our tool, Sarah! With the smallest nuclei, it is possible that nucleus X has non-zero probability for a bunch of voxels, but does not win any of those voxels (i.e., there’s another label with larger probability). If that happens, the volume will be non-zero (as you have some volume coming from the non-zero probabilities) but no single label will be labeled as X.
Does this make sense?
Cheers,
/Eugenio
--
Juan Eugenio Iglesias
http://secure-web.cisco.com/1vPzx65W4UOXdNeMIdjQo59tCy-oDZ1X15-uLq-KhzdF-WND...
From: freesurfer-bounces@nmr.mgh.harvard.edu on behalf of Sarah Skinner seskinner@wisc.edu Reply-To: Freesurfer support list Freesurfer@nmr.mgh.harvard.edu Date: Friday, January 13, 2023 at 11:46 AM To: "Freesurfer@nmr.mgh.harvard.edu" Freesurfer@nmr.mgh.harvard.edu Subject: [Freesurfer] Hippocampal/Amygdala subfield segmentation error - Anterior Amygdala Area missing in FreeView
External Email - Use Caution
Hi FreeSurfer Experts,
I'm attempting to QC the hippocampal/amygdala segmentation in FreeView. When looking at ?h.hippoAmygLabels-T1.v21.FS60.mgz I can see that a few subjects are either missing a mask of the anterior amygdaloid area (AAA) for a single hemisphere or the AAA is extremely small. However, in both circumstances the volumetric value for the AAA in the output (?h.amygNucVolumes-T1.v21.txt) looks normal. In the cases where no AAA mask is visible, if you check the box "Show existing labels only", the label "7010 Anterior-amygdaloid-area" does not even appear for that hemisphere.
AAA is the only structure where this is occurring. There are other structures (such as the medial nucleus) that appear small in FreeView, but still have a normal volumetric value. However, no other structure is completely missing.
Note: In all of these cases aseg looks fine. And T1s were processed using freesurfer-Darwin-OSX-stable-pub-v6.0.0-2beb96c, but we used freesurfer-darwin-macOS-7.1.0-20200511-813297b to run with hippocampal/amygdala segmentation (segmentHA_T1.sh).
Thank you for you time and help,
Sarah Skinner
Research Specialist University of Wisconsin - Madison
freesurfer@nmr.mgh.harvard.edu
-
Iglesias Gonzalez, Juan E. -
Sarah Skinner