Hey Allison,
Thanks for you help. What do you mean by check your talairach? I'll hold off on using a different set of dicoms, and make that my last resort since it takes a day to complete. Thanks, m ________________________________________ From: Allison Stevens Player [astevens@nmr.mgh.harvard.edu] Sent: Monday, May 14, 2012 3:16 PM To: Borzello, Mia Cc: freesurfer@nmr.mgh.harvard.edu Subject: RE: [Freesurfer] "cannot allocate memory" issue
Hi Mia, You may want to check your talairach to make sure that looks okay. If that's not causing the problem, you can edit the wm.mgz to remove the cerebellum and then rerun and this should fix the problem. Allison
On Mon, 14 May 2012, Borzello, Mia wrote:
It says that it couldn't open rh.orig.nofix and lh.orig.nofix. On the popup image, it doesn't look like the skull is present, but it does look like the cerebellum is. ________________________________________ From: Allison Stevens Player [astevens@nmr.mgh.harvard.edu] Sent: Monday, May 14, 2012 1:28 PM To: Borzello, Mia Cc: freesurfer@nmr.mgh.harvard.edu Subject: RE: [Freesurfer] "cannot allocate memory" issue
You can do: tkmedit subjid filled.mgz rh.orig.nofix -aux-surface lh.orig.nofix
When you downloaded FreeSurfer, you would have also received the subject "bert" who is a good example to compare to. He will likely be located in: $FREESURFER_HOME/subjects/bert
If you do the same command above for bert: tkmedit bert filled.mgz rh.orig.nofix -aux-surface lh.orig.nofix
You'll see what the filled.mgz should look like. Allison
On Mon, 14 May 2012, Borzello, Mia wrote:
How do I check this? (Apologies for my ignorance.) So I might need to just use a different set of dicoms then?
thanks, Mia ________________________________________ From: Allison Stevens Player [astevens@nmr.mgh.harvard.edu] Sent: Monday, May 14, 2012 1:02 PM To: Borzello, Mia Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] "cannot allocate memory" issue
It looks like there is too huge a topology error for it to continue:
CORRECTING DEFECT 23 (vertices=143258, convex hull=19937) Excessive topologic defect encountered: could not allocate -729782272 edges for retessellation Cannot allocate memory
I would suggested looking at the filled.mgz to make sure skull and/or cerebellum are not included. Allison
On Mon, 14 May 2012, Borzello, Mia wrote:
Hi all,
I was wondering if you could help me figure out why the recon isn't running to completion. I've attached the log file (recon-all.log), but it's unclear to me what I should do to fix this problem. I tried to compute the lh.pial, but I guess it was never made in the recon.
Thanks so much, Mia