External Email - Use Caution        

Dear Bram,

It looks like the -bigventricles flag fixed my problem with my specific subject. Not only, this took into account large ventricles but also corrected grey matter segmentation from extra Dura tissue that was erroneously segmented in GM.

I understand the first point of amelioration allowed by the flag but don't understand where the second point come from ?

If I used the -bigventricles flag for all my atrophied and large ventricles patients and not for my healthy patients, wouldn't it introduce a bias when further statistically comparing groups between them ?

Best,
Matthieu

2018-09-14 15:35 GMT+00:00 Diamond, Bram Ryder <BRDIAMOND@mgh.harvard.edu>:
If you haven't already, you may want to run the subject through recon-all with the -bigventricles flag since it looks like the lateral ventricle was significantly mislabeled in the aseg. If using the -bigventricles flag doesn't fix your problem, you can edit the aseg.mgz manually by following the directions in this tutorial. Then run the following command (again, substituting <subj_id> for your subject's id):

recon-all -autorecon2-noaseg -autorecon3 -subjid <subj_id>

I would try this before doing the wm.mgz edits I suggested in my previous email.

Best,
Bram
From: freesurfer-bounces@nmr.mgh.harvard.edu <freesurfer-bounces@nmr.mgh.harvard.edu> on behalf of Diamond, Bram Ryder <BRDIAMOND@mgh.harvard.edu>
Sent: Friday, September 14, 2018 11:07:42 AM
To: matthieuvanhoutte@gmail.com
Cc: freesurfer@nmr.mgh.harvard.edu; astevens@nmr.mgh.harvard.edu
Subject: [Freesurfer] Troubles to determine the type of recon editing needed
 

Hi Matthieu,


I've taken a look through the files you shared with us and I see the poor surfaces in the posterior right hemisphere you were referring to in your message. It looks like your subject has a combination of abnormally large ventricles and significant wm abnormalities, so I'm surprised FreeSurfer did as well as it did.


As for the recon editing - I would recommend editing the wm.mgz to more accurately represent the wm from slice 84 to 39. You can take a look at the white matter edits tutorial for details on how to do that. Then run the following command (substituting <subj_id> for your subject's id):


recon-all -autorecon2-wm -autorecon3 -subjid <subj_id>


The surface reconstruction may also benefit from labeling the right lateral ventricle in the wm.mgz (as an intensity of 250) - but I'm not certain since you didn't send us the surfaces for the left hemisphere. Before you do that, edit the wm.mgz as explained in the tutorial and tell us how that goes.


All the best,

Bram


Bram R. Diamond, BSc
Research Technician II
Laboratory for Computational Neuroimaging
Martinos Center for Biomedical Imaging
Massachusetts General Hospital
(p): 617-726-6598

The information in this e-mail is intended only for the person to whom it is
addressed. If you believe this e-mail was sent to you in error and the e-mail
contains patient information, please contact the Partners Compliance HelpLine at
http://www.partners.org/complianceline . If the e-mail was sent to you in error
but does not contain patient information, please contact the sender and properly
dispose of the e-mail.