Hi Derin,
the bad talairach is probably messing everything up. You can either correct it manually with tkregister2, or remove/rename the talairach.xfm and talairach.auto.xfm use recon-all with -notalairach so it won't be rebuilt. Nick: is there a flag to ignore the tal xform even if it's there? Something like -badtalairach?
cheers, Bruce
On Mon, 2 Apr 2007, Derin Cobia wrote:
Two questions for two different subjects during autorecon1:
- The initial skull strip on this subject left too much extra-cerebral
material. Subsequent attempts at using the wsatlas or manual adjustment of the wsthresh have resulted in this error:
*************************WATERSHED************************** preflooding height equal to 25 percent Sorting... T1-weighted MRI image modification of the preflooding height to 15 percent Count how many 110 voxels are present : 1960516
Find the largest 110-component...mriSegmentReallocateVoxels: couldnot alloc 1048381 voxels for sno 8343 Cannot allocate memory done And identify it as the main brain basin...Segmentation fault Linux rainmaker 2.6.9-42.0.3.ELsmp #1 SMP Fri Oct 6 06:21:39 CDT 2006 i686 i686 i386 GNU/Linux
recon-all exited with ERRORS at Mon Apr 2 15:02:05 CDT 2007 #--------------------------------------------------------------
Are there too many voxels? I've looked, but couldn't find any information on how to fix this one.
- This subject continues to fail with errors during intensity
normalization. Here is the output:
#-------------------------------------------- #@# Intensity Normalization Mon Apr 2 15:46:47 EDT 2007 /disk/conte_raid1/freesurfer/conte_mprage/021219_5734/mri
mri_normalize -g 1 nu.mgz T1.mgz
using max gradient = 1.000 reading from nu.mgz... normalizing image... talairach transform 0.830 -0.117 -0.004 -3.656; 0.116 0.833 -0.323 -26.577; 0.004 0.028 0.075 84.152; 0.000 0.000 0.000 1.000; INFO: Modifying talairach volume c_(r,a,s) based on average_305 MRInormalize: could not find any valid peaks mri_normalize: normalization failed Linux eli 2.6.9-34.0.1.ELsmp #1 SMP Wed May 17 16:59:36 EDT 2006 x86_64 x86_64 x86_64 GNU/Linux
recon-all exited with ERRORS at Mon Apr 2 15:46:49 EDT 2007 #---------------------------------------------
I've checked the talairach.xfm and it is bad. It seems I remember someone on the list saying that a -nold flag could fix this, but I couldn't find any info on the wiki about it. Any suggestions? Can I still salvage this subject?
Thanks for all of your help!
-Derin _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer