Hi Thomas
when you say errors in the hippocampus do you mean the aseg or the surfaces? If the surfaces, don't worry about them as they are not used in the hippocampal region. You are probably connecting the temporal lobe to the main body of the white matter creating a giant defect.
cheers Bruce On Mon, 20 Jan 2014, Thomas Fink wrote:
Dear Freesurfer experts,
We are using FS 5.1 and have several participants with errors in the segmentation of the hippocampus and other regions of the cortex. So far we did corrections to the hippocampus but in 10 out of 13 participants we got error messages listed below:
Seven participants:
CORRECTING DEFECT 0 (vertices=74686, convex hull=1605)
Excessive topologic defect encountered: could not allocate 390335770 edges for retessellation
Cannot allocate memory
Darwin NPSY.local 12.4.0 Darwin Kernel Version 12.4.0: Wed May 1 17:57:12 PDT 2013; root:xnu-2050.24.15~1/RELEASE_X86_64 x86_64
-----
One participant:
mhtVoxelList_Add(189, 240, 290, 260205): complete list too big!
mhtVoxelList_Add(189, 240, 290, 260205): complete list too big!
No such file or directory
marking ambiguous vertices...
Too many intersected faces for face 255461 (126442, 128896, 128873)
No such file or directory
Darwin NPSY.local 12.4.0 Darwin Kernel Version 12.4.0: Wed May 1 17:57:12 PDT 2013; root:xnu-2050.24.15~1/RELEASE_X86_64 x86_64
One participant:
-----
CORRECTING DEFECT 14 (vertices=33317, convex hull=1285)
mris_fix_topology(26738) malloc: *** mmap(size=1212047360) failed (error code=12)
*** error: can't allocate region
*** set a breakpoint in malloc_error_break to debug
Bus error
Darwin NPSY.local 12.4.0 Darwin Kernel Version 12.4.0: Wed May 1 17:57:12 PDT 2013; root:xnu-2050.24.15~1/RELEASE_X86_64 x86_64
One participant:
Correction of the Topology
Finding true center and radius of Spherical Surface...done
Surface centered at (0,0,0) with radius 100.0 in 11 iterations
marking ambiguous vertices...
414500 ambiguous faces found in tessellation
segmenting defects...
Bus error
Darwin NPSY.local 12.4.0 Darwin Kernel Version 12.4.0: Wed May 1 17:57:12 PDT 2013; root:xnu-2050.24.15~1/RELEASE_X86_64 x86_64
The error log says in all cases that there is something wrong with the cerebellum. When we open the ?h inflated.nofix file then we see the cerebellum added (see attached file) in the hemisphere we made the edits.
We have no clue if or what we are doing wrong and here are the steps we did:
- open the files (tkmedit subject norm.mgz -aseg.mgz
- Tool -> Configure Segmentation Brush info -> select the hippocampus -> add
voxels
File -> save segmentation
run recon with "recon-all -autorecon2 -noaseg –autorecon3 -subjid subject
-nuintensitycor-3T"
So far we had no problems with our subjects for all the previous steps (recon after wm, pial, .... edits) and it did run smoothly.
Any help would be greatly appreciated.
Thomas Fink Mail: Thomas.Fink@psy.lmu.de