I guess it is detecting the fact that gcuts failed. Nick: is there any way to disable the failure detection so Jacquie can see what it did? Bruce
On Thu, 4 Aug 2011, Jacquelynn Nicosia Copeland wrote:
Hi Khoa,
Thanks for your response. The extra dura/skull is being included in some parts of the pial surface, particularly around the superior surfaces of the brain (gray matter and cortical regions that we are interested in are being affected). Thus, manual editing by erasing dura/skull has been taking a long time (hours?) to do in on one participant.
That's why I tried gcut on the original image to see if it was much different, but I got the message I previously posted, where it appeared that gcut did not work. I don't really care about cutting off the cerebellum, as long as it takes off the extra dura/skull affecting the pial surface and thus decreasing the amount of manual edits I need to make.
I will try to the additional skull strip as you suggested. But why do you think that gcut isn't running and I am getting the message I included in my earlier post? I was really hoping gcut would be my savior rather than making lots of edits. In addition, adjusting the watershed parameters hasn't really helped either. Is there no saving grace?
Thanks, Jacquie
Jacquelynn N. Copeland, M.A. Doctoral Candidate Medical/Clinical Psychology Program Department of Psychology University of Alabama at Birmingham Email:jcopeland@uab.edu ________________________________________ From: Khoa Nguyen [khoa@nmr.mgh.harvard.edu] Sent: Thursday, August 04, 2011 6:38 AM To: Jacquelynn Nicosia Copeland Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] skull stripping and gcut error
Hi,
How much dura/skull is left on your subjects that you want to remove? Sometimes gcut takes away more than just dura, especially in the cerebellum region so that might be why it is much smaller.
If the surfaces are not including the extra dura/skull, then it is not necessary that you remove them. If you want, you can try just running "recon-all -skullstrip -s subjid" again instead. That might improve it.
-Khoa
On Wed, 3 Aug 2011, Jacquelynn Nicosia Copeland wrote:
Hi Freesurfers:
I am trying to minimize the amount of skull stripping manual edits I need to make on Freesurfer outputs due to remaining dura. I've been trying both: adjusting the watershed parameters and using gcut.
However, I noticed the output with and without gcut for the same participant were not different. So I went to the log file, and when I searched for gcut it says:
" rm -f brainmask.gcuts.mgz
mri_gcut -110 -mult brainmask.auto.mgz T1.mgz brainmask.auto.mgz brainmask.gcuts.mgz
INFO: Care must be taken to thoroughly inspect your data when using mri_gcut. In particular, inspect the edges of gm and cerebellum for over-aggressive cutting. Add -segmentation brainmask.gcuts.mgz to the tkmedit command string to view the voxels which gcut has removed.
reading mask... white mean: 107.529232 threshold set to: 107.529232*0.400000=43.011693 calculating weights... doing mincut... now doing maxflow, be patient... g-cut done! post-processing... post-processing done! masking... ** Gcutted brain is much smaller than the mask! ** Using the mask as the output instead!
cp brainmask.auto.mgz brainmask.mgz "
Thus, it does not appear to be using the brainmask with gcut edits. I added the "-segmentation brainmask.gcuts.mgz" to the tkmedit command and got an error basically saying that the file does not exist. This is the case for many of our participants. Why is this occurring? Isn't the purpose of gcut is to make the brainmask smaller (b/c it has less dura in it)?
Also, any other suggestions for reducing remaining dura/skull?
Thanks for any help you can provide, Jacquie
Jacquelynn N. Copeland, M.A. Doctoral Candidate Medical/Clinical Psychology Program Department of Psychology University of Alabama at Birmingham Email:jcopeland@uab.edu _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
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.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer