Hi Bruce and all,

Thank you for your help. Bill Monaco emailed and let me know that this happened to him as well. He suggested running the '-autorecon2 -autorecon3' flags on this subject (as opposed to the 'recon-all -all'), and this worked for me!

However, here is the end of the recon-all.log file, where the CA Reg error occured:

pct change decreased switching integration type to fixed (done=0) pct change decreased pct change < tol 0.025, nsmall = 1 of 2 pct change decreased pct change decreased pct change decreased pct change decreased pct change decreased pct change decreased pct change decreased pct change decreased pct change decreased pct change decreased pct change decreased pct change < tol 0.025, nsmall = 1 of 2 pct change decreased pct change decreased pct change < tol 0.025, nsmall = 1 of 2 pct change decreased pct change < tol 0.025, nsmall = 1 of 2 pct change < tol 0.025, nsmall = 2 of 2 switching integration type to fixed (done=1) pct change decreased pct change < tol 0.025, nsmall = 1 of 2 pct change < tol 0.025, nsmall = 2 of 2 setting smoothness coefficient to 0.40 blurring input image with Gaussian with sigma=2.000... rms increased - undoing step... blurring input image with Gaussian with sigma=0.500... pct change decreased rms increased - undoing step... setting smoothness coefficient to 1.00 blurring input image with Gaussian with sigma=2.000... pct change decreased switching integration type to fixed (done=0) rms increased - undoing step... pct change < tol 0.025, nsmall = 1 of 2 blurring input image with Gaussian with sigma=0.500... pct change decreased swSegmentation fault ERROR: mri_ca_register with non-zero status Linux node0052 2.6.9-42.0.3.ELsmp #1 SMP Fri Oct 6 06:21:39 CDT 2006 i686 athlon i386 GNU/Linux

recon-all exited with ERRORS at Wed Oct 24 07:03:49 EDT 2007

Thank you, Caroline

On Thu, 6 Dec 2007, Bruce Fischl wrote:

Hi Caroline,

can you send us the end of the recon-all.log when it fails?

Bruce On Thu, 6 Dec 2007, Caroline Chapman wrote:

...

Hello,

I am having trouble with one subject during the recon-all process. It has failed twice at CA Reg:

status file for recon-all Thu Nov 8 15:11:20 EST 2007 #@# MotionCor Thu Nov 8 15:11:20 EST 2007 #@# Nu Intensity Correction Thu Nov 8 15:18:50 EST 2007 #@# Talairach Thu Nov 8 15:21:26 EST 2007 #@# Intensity Normalization Thu Nov 8 15:23:01 EST 2007 #@# Skull Stripping Thu Nov 8 15:30:05 EST 2007 #@# EM Registration Thu Nov 8 15:33:05 EST 2007 #@# CA Normalize Thu Nov 8 15:42:54 EST 2007 #@# CA Reg Thu Nov 8 15:45:14 EST 2007 Linux node0052 2.6.9-42.0.3.ELsmp #1 SMP Fri Oct 6 06:21:39 CDT 2006 i686 athlon i386 GNU/Linux

The brainmask, orig, and talairach look okay. Any ideas why this might be failing?

Thanks! Caroline

Hi,

I am trying to register the spheres to the template. The current version of mris_register may have some nonlinear local deformation of the gyral and sulcal structure. Anyway I can specify the "rigid alignment" only? I didn't find the option for linear alignment only after checking the options of mris_register.

I am using FreeSurfer v4.0.1. Thank you for your help.

XJ Kang

mris_register [options] <input surface> <average surface> <output surface>

Options:

-norot : disable initial rigid alignment -nosulc : disable initial sulc alignment -curv : use smoothwm curvature for final alignment -jacobian <fname> : write-out jacobian overlay data to fname -dist <num> : specify distance term -l <label file> <atlas (*.gcs)> <label name> : this option will specify a manual label to align with atlas label <label name> -addframe <which_field> <where_in_atlas> <l_corr> <l_pcorr> -overlay <surfvals> <navgs> : subject/labels/hemi.surfvals -overlay-dir <dir> : subject/dir/hemi.surfvals -1 : target specifies a subject's surface, not a template file

Hi,

I ran "make_average_subject" to get averaged volume and surface. The average inflated surface looks like "half-inflated", not like individual's fully inflated smooth surface. I am not sure if it is depends on the subject pool. Any way to make it fully inflated? Thanks.

XJ Kang

yes, you can give mris_inflate a -n <iter> option, where <iter> is 10 by default. The average surfaces don't always inflate the same as the individuals, but it's only for display so we've never tried very hard to fix it.

cheers, Bruce

On Fri, 25 Jan 2008, Nick Schmansky wrote:

You could use mris_inflate to inflate it more if you prefer.

On Fri, 2008-01-25 at 14:10 -0800, XJ Kang wrote:

...

Hi,

I ran "make_average_subject" to get averaged volume and surface. The average inflated surface looks like "half-inflated", not like individual's fully inflated smooth surface. I am not sure if it is depends on the subject pool. Any way to make it fully inflated? Thanks.

XJ Kang _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer