Hi Chad,
Why do you say the Segmentation step overwrites your WM edits?  We've just run WM edits under FS 5.3 very recently and it worked fine.  Also, a glance at the recon-all script indicates that the Segmentation step is testing for the presence of existing edits.

cheers,
-MH
-- 
Michael Harms, Ph.D.
-----------------------------------------------------------
Conte Center for the Neuroscience of Mental Disorders
Washington University School of Medicine
Department of Psychiatry, Box 8134
660 South Euclid Ave. Tel: 314-747-6173
St. Louis, MO  63110 Email: mharms@wustl.edu

From: <Johnson>, Chad P <Chad.P.Johnson@uth.tmc.edu>
Date: Tuesday, February 4, 2014 10:27 AM
To: "Freesurfer@nmr.mgh.harvard.edu" <Freesurfer@nmr.mgh.harvard.edu>
Subject: [Freesurfer] Autorecon2 Pipeline Problem

Good morning,

 

I am new to the e-mail list and would like to request help with a problem that has arisen when running recon-all.  As such, I apologize if this question has been posed previously.  We are running Freesurfer version 5.3.0 on Red Hat Enterprise Linux Release 6.3 using the freesurfer-Linux-centos6_x86_64-stable-pub-v5.3.0 stable release of Freesurfer. 

 

Following white matter edits, the invocation “recon-all –autorecon2-wm –s <subjid>” begins at the Normalization 2 step, rather than the Fill step as stated in the processing stages pipeline.  This is problematic as the following segmentation step overwrites the edited wm.mgz file with new automated information.  Review of the output also indicates that the wm.mgz is being regressed back to the pre-edited state.  A check of the recon-all lib file revealed the following.

 

recon-all lib file source code from version 5.3.0 (lines 5465 through 5482):

 

    case "-autorecon2-cp":

    case "-autorecon2-noaseg":

    case "-autorecon2-wm":

      set DoNormalization2 = 1;

      set DoSegmentation   = 1;

      set DoFill           = 1;

      set DoTessellate     = 1;

      set DoSmooth1        = 1;

      set DoInflate1       = 1;

      set DoQSphere        = 1;

      set DoFix            = 1;

      set DoMaskBFS        = 1;

      set DoWhiteSurfs     = 1;

      set DoSmooth2        = 1;

      set DoInflate2       = 1;

      set DoSegStats       = 1;

      set DoCurvStats      = 1;

      breaksw

 

Where it appears that –cp, –noaseg, and –wm are all aliases of the same command.  This differs significantly from the previous release we used in the lab of 5.1.0.

 

recon-all lib file source code from version 5.1.0 (lines 4806 through 4819):

 

    case "-autorecon2-wm":

      set DoFill           = 1;

      set DoTessellate     = 1;

      set DoSmooth1        = 1;

      set DoInflate1       = 1;

      set DoQSphere        = 1;

      set DoFix            = 1;

      set DoMaskBFS        = 1;

      set DoWhiteSurfs     = 1;

      set DoSmooth2        = 1;

      set DoInflate2       = 1;

      set DoSegStats       = 1;

      set DoCurvStats      = 1;

      breaksw

 

I assume that I can modify my pipeline with some –no commands to keep the DoNormalization2 and DoSegmentation steps from running but I was wondering if there was a reason for the change between versions which would argue against starting at the fill stage following white matter edits.  I can provide any additional information if needed to help solve this problem and any information that can be provided regarding this would be most helpful.  Thanks.

 

 

Chad P. Johnson, Ph.D.

Postdoctoral Fellow

University of Texas Health Science Center

 

The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.