I'm having problems with recon-all. It used to work but recently, every time I've used it, I get a nu_correct error. I don't think we have had any freesurfer upgrades since the last time recon-all worked. Here's the log. Anyone know what's wrong?
ip07:mperry[12] setenv SUBJECTS_DIR <path/to/subjects/dirs> ip07:mperry[13] setenv FUNCTIONALS_DIR <path/to/subjects/dirs> ip07:mperry[14] recon-all -stage1 -subjid subject_ID ######################################## program versions used $Id: recon-all-nmr,v 1.50 2004/08/31 18:05:08 tosa Exp $ $Id: mri_motion_correct2,v 1.8 2004/10/04 22:52:12 greve Exp $ mri_convert --version $ $Header: /software/source/minc/cvsroot/minc/progs/rawtominc/rawtominc.c,v 6.12 2 003/11/14 16:52:24 stever Exp $ minctracc.c $Header: /software/source/registration/mni_autoreg/minctracc/Main/minctracc.c,v 96.5.2.1 2003/07/05 01:58:08 stever Exp $ $Header: /software/source/minc/cvsroot/minc/progs/mincresample/mincresample.c,v 6.12 2003/09/18 15:01:33 bert Exp $ $Header: /software/source/minc/cvsroot/minc/progs/mincaverage/mincaverage.c,v 6. 4 2001/04/24 13:38:42 neelin Exp $ $Id: talairach2,v 1.7 2004/08/31 18:07:37 tosa Exp $ mri_convert --version $ # $Id: mritotal.in,v 1.11.2.1 2003/11/17 06:01:15 stever Exp $ # $Id: nu_estimate.in,v 1.1 2003/04/16 14:29:34 bert Exp $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ ####################################### INFO: SUBJECTS_DIR is <path/to/subjects/dirs> --------------------------------------------- MotionCor Tue Mar 15 16:03:21 PST 2005 <path/to/subjects/dirs>/<subject_ID>/mri/orig mri_motion_correct2 -o . -i 001 -i 002 tmpdir is ./tmp-mri_motion_correct2-23141 -------------------------------------------------------------- mri_motion_correct2 logfile is ./mri_motion_correct2.log -------------------------------------------------------------- ----------------------------------------- Converting 001 Tue Mar 15 16:03:21 PST 2005 mri_convert 001 <path/to/subjects/dirs>/<subject_ID>/mri/orig/tmp-mri_mo tion_correct2-23141/cor-1.mnc reading from 001... TR=2730.00, TE=3.31, TI=1000.00, flip angle=7.00 i_ras = (-1, 0, 0) j_ras = (0, -0, -1) k_ras = (0, 1, -0) writing to <path/to/subjects/dirs>/<subject_ID>/mri/orig/tmp-mri_motion_ correct2-23141/cor-1.mnc... ----------------------------------------- Converting 002 Tue Mar 15 16:03:29 PST 2005 mri_convert 002 <path/to/subjects/dirs>/<subject_ID>/mri/orig/tmp-mri_mo tion_correct2-23141/cor-2.mnc reading from 002... TR=2730.00, TE=3.31, TI=1000.00, flip angle=7.00 i_ras = (-1, 0, 0) j_ras = (0, -0, -1) k_ras = (0, 1, -0) writing to <path/to/subjects/dirs>/<subject_ID>/mri/orig/tmp-mri_motion_ correct2-23141/cor-2.mnc... ----------------------------------------- Motion Correcting <path/to/subjects/dirs>/<subject_ID>/mri/orig/tmp-mri_ motion_correct2-23141/cor-2.mnc Tue Mar 15 16:03:36 PST 2005 COG of v1: 2.191772 14.916920 -47.842663 COG of v2: 2.203371 14.932534 -47.776924 [trans] = 0.011599 0.015615 0.065739 Initial objective function val = 0.01807109 Final objective function value = 0.00892547 ----------------------------------------- Resampling <path/to/subjects/dirs>/<subject_ID>/mri/orig/tmp-mri_motion_ correct2-23141/cor-2.mnc Tue Mar 15 16:04:30 PST 2005 Transforming slices:............................................................ ................................................................................ ................................................................................ ....................................Done ----------------------------------------- Averaging Tue Mar 15 16:04:37 PST 2005 Processing:..................................................................... ................................................................................ ................................................................................ ...........................Done ----------------------------------------- Converting Average to output Tue Mar 15 16:04:41 PST 2005 mri_convert <path/to/subjects/dirs>/<subject_ID>/mri/orig/tmp-mri_motion _correct2-23141/avgvol.mnc . -tr 2730 -te 3.31 -TI 1000 -flip_angle 0.122173 reading from <path/to/subjects/dirs>/<subject_ID>/mri/orig/tmp-mri_motio n_correct2-23141/avgvol.mnc... TR=0.00, TE=0.00, TI=0.00, flip angle=0.00 i_ras = (-1, 0, 0) j_ras = (0, -0, -1) k_ras = (0, 1, -0) Original Data has (1, 1, 1) mm size and (256, 256, 256) voxels. Data is conformed to 1 mm size and 256 voxels for all directions writing to .... Started at: Tue Mar 15 16:03:21 PST 2005 Ended at: Tue Mar 15 16:04:46 PST 2005 mri_motion_correct2: done --------------------------------------------- Nu Intensity Correction Tue Mar 15 16:04:46 PST 2005 mri_convert orig nu0.mnc mri_convert orig nu0.mnc reading from orig... TR=2730.00, TE=3.31, TI=1000.00, flip angle=7.00 i_ras = (-1, 0, 0) j_ras = (0, -0, -1) k_ras = (0, 1, -0) writing to nu0.mnc... nu_correct -clobber nu0.mnc nu1.mnc [mperry@ip07:<path/to/subjects/dirs>/<subject_ID>/mri/] [2005-03-15 16:0 4:54] running: /usr/pubsw/packages/mni/1.2/bin/nu_estimate_np_and_em -parzen -log -sharpen 0. 15 0.01 -iterations 50 -stop 0.001 -shrink 4 -auto_mask -nonotify -b_spline 1 -d istance 200 -quiet -execute -clobber -nokeeptmp -tmpdir /usr/tmp/nu_correct_2326 4/ nu0.mnc nu1.imp
spawn: exec of sharpen_volume failed: No such file or directory nu_estimate_np_and_em: crashed while running sharpen_volume (termination status= 65280) nu_correct: crashed while running nu_estimate_np_and_em (termination status=6528 0) ERROR: nu_correct ------------------------------------------ Filesystem 1K-blocks Used Available Use% Mounted on
693727248 307742336 350745536 47% /space/aliens/1 ------------------------------------------ recon-all exited with errors at Tue Mar 15 16:04:55 PST 2005
Hi. I am going to need a sample of points distributed equally over the cortical surface, with a given spacing between them.
I noticed that freesurfer has the capability to do this, but I am wondering HOW it is done, to see if indeed it does what I expect. (tools - decimate)
In particular, I would like a "surface distance" measure of the spacing, not a 3D euclidean distance as a measure of the spacing, or some other measure.
I did a little test, and it was strange. I decimated with a distance of 6, and with 10. I then asked matlab to find the minimum distance between vertices for all vertices in the decimated set, and got mean and median values of those.
Here is the weird part. For the spacing of "6", the mean and median of the minimum distances were 8 and 9 or so. For the spacing of "10", the mean and median of the min distances were 15 and 13.
Realizing that my technique of getting the minimum euclidean distance is WRONG, but it should provide a simple approximation. That is, my measure should be SMALLER than the surface distance measure. Therefore, it doesn't make sense to me that the decimation is doing what I was hoping it does, because for surface spacings of 6, I expect euclidean distances to be less than 6, and for surface spacings of 10, I expect euclidean distances to be less than 10.
Thoughts from freesurfing fellows?
freesurfer@nmr.mgh.harvard.edu
-
Daniel Goldenholz -
Michele Perry