13 Oct
2015
13 Oct
'15
12:17 p.m.
Hi Anthony
feel free to email me the .gca and I'll take a look
cheers Bruce On Tue, 13 Oct 2015, Anthony Dick wrote:
Hello Bruce,
I used the rebuild_gca_atlas.csh script (attached). I only commented out the cluster configuration options. Otherwise the script is unchanged.
The process fails in the same place when we use one subject.
Thanks!
Anthony
On 10/13/15 12:07 PM, Bruce Fischl wrote: and how did you create the gca? On Tue, 13 Oct 2015, Anthony Dick wrote:
Hello Bruce, I will try a single subject. When I run the following command line (i.e., right out of recon-all) it works: mri_em_register -uns 3 -mask brainmask.mgz nu.mgz $FREESURFER_HOME/average/RB_all_2014-08-21.gca transforms/talairach.lta When I replace the Freesurfer-shipped .gca with the script created one I get the following: mri_em_register -uns 3 -mask brainmask.mgz nu.mgz /<path>/average/RB_one_2015-10-13.gca transforms/talairach.lta setting unknown_nbr_spacing = 3 using MR volume brainmask.mgz to mask input volume... == Number of threads available to mri_em_register for OpenMP = 16 == reading 1 input volumes... logging results to talairach.log reading '<path>/average/RB_one_2015-10-13.gca'... average std = 9.8 using min determinant for regularization = 9.7 0 singular and 426 ill-conditioned covariance matrices regularized reading 'nu.mgz'... freeing gibbs priors...done. bounding unknown intensity as < 18.3 or > 695.1 Segmentation fault So it appears to be a problem with the gca creation? However, mri_info on both RB_all_2014-08-21.gca and RB_one_2015-10-13.gca look identical. Anthony On 10/13/15 11:54 AM, Bruce Fischl wrote: HI Anthony what is the full command line? Does it happen if you only train on a single subject instead of 23? Bruce On Tue, 13 Oct 2015, Anthony Dick wrote: Hello Bruce, We have tried running the atlas creation on newly freesurfed brains with the latest FS version and keep getting the same error. We have essentially replicated the error we get on our own manually edited volumes by copying the aparc+aseg.mgz and renaming it seg_edited.mgz. Thus, it is unlikely to be a problem with our manually edited brains. This results in the following error at the same place (running the rebuild_gca_atlas.csh script): == Number of threads available to mri_em_register for OpenMP = 16 == reading 23 input volumes... logging results to talairach_one.log reading '<path>/average/RB_one_2015-10-13.gca'... average std = 9.8 using min determinant for regularization = 9.7 0 singular and 426 ill-conditioned covariance matrices regularized reading '<path>/009/mri/nu.mgz'... reading '<path>/average/RB_one_2015-10-13.gca'... reading frame -1 of gca INFO: MRImask() using MRImaskDifferentGeometry() INFO: MRImask() using MRImaskDifferentGeometry() INFO: MRImask() using MRImaskDifferentGeometry() INFO: MRImask() using MRImaskDifferentGeometry() INFO: MRImask() using MRImaskDifferentGeometry() reading '<path>/009/mri/transforms/talairach_one.lta'... mri_read(): couldn't determine type of file <path>/009/mri/transforms/talairach_one.lta mri_em_register: could not open input volume <path>/009/mri/transforms/talairach_one.lta. Perhaps the FS folks could try to replicate this error? You don't need my brains, just a couple of FS'd brains with a file called seg_edited.mgz. If you don't get the error, then maybe it is a problem with our local settings (although we get this error on multiple different computers). Thanks for any help! Anthony On 7/20/15 2:30 PM, Anthony Dick wrote: Hello, I was wondering if these got uploaded correctly. Please let me know if I made a mistake and thanks, Anthony On 7/9/15 1:41 PM, Bruce Fischl wrote: sure On Thu, 9 Jul 2015, Anthony Dick wrote: Hello Bruce, I am still getting this error after re-running autorecon1 in the newest version of FS. Can I upload two brains and see what is the issue? I'm a bit stumped on this one. Anthony On 5/26/15 12:21 PM, Bruce Fischl wrote: Hi Anthony sorry, I know you've posted this before. What version are you running? And what type is your labeled data? That is, are they uchar? int? At some point this used to not work with non-uchar data, but I think I fixed it a while ago. YOu might try downloading a beta of V6 and see if it fixes your problem cheers Bruce On Tue, 26 May 2015, Anthony Dick wrote: Hello, I have been stumped by this error I keep getting using the rebuild_gca_atlas.csh script. I am trying to build a cerebellar atlas, so I am wondering if the problem is with the brains we parcellated, or with my installation etc. The brains consist of manually segmented cerebellums from the original T1. The T1 is run through Freesurfer, and the ROIs are made in Freeview. The edited file is called seg_edited.mgz. Things run smoothly up until this error: using MR volume /<path>/subjects/4/mri/brain.mgz to mask input volume... reading 23 input volumes... logging results to talairach_one.log reading 'subjects/average/RB_one_2015-05-14.gca'... gcaAllocMax: node dims 32 32 32 gcaAllocMax: prior dims 128 128 128 gcaAllocMax: max_labels 0 average std = 8.4 using min determinant for regularization = 7.1 0 singular and 119 ill-conditioned covariance matrices regularized reading '/<path>/subjects/4/mri/nu.mgz'... reading '/<path>/subjects/average/RB_one_2015-05-14.gca'... gcaAllocMax: node dims 32 32 32 gcaAllocMax: prior dims 128 128 128 gcaAllocMax: max_labels 0 reading frame -1 of gca MRIcopyFrame: src and dst must be same type reading '/<path>/subjects/4/mri/transforms/talairach_one.lta'... mri_read(): couldn't determine type of file /<path>/subjects/4/mri/transforms/talairach_one.lta mri_em_register: could not open input volume /<path>/subjects/4/mri/transforms/talairach_one.lta. _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail. _______________________________________________ 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
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mai l contains patient information, please contact the Partners Compliance HelpLin e at http://www.partners.org/complianceline . If the e-mail was sent to you in er ror but does not contain patient information, please contact the sender and prop erly dispose of the e-mail.