Sorry for the delay in responding, I wanted to try every permutation that I could think of. I updated the tracula binaries for snow leopard (https://mail.nmr.mgh.harvard.edu/pipermail//freesurfer/2011-August/019946.ht...) and that gives me some of the same, and some different issues that may help. First, if the bvec file is in the correct format (3 columns), then it copies the file correctly and trac-all continues smoothly. If the file is in the 3 row format (needing to be transposed), then tracula will only copy the first 9 numbers out of it giving the error of not having the correct number of bvecs to bvals. So that seems fixable, I just need to reorient all of our bval and bvec files into the correct format.
I'm still a bit unsure why tracula won't create the bvec and bval files during the mri_convert stage. I have been using dcm2nii to create these two files but then continue to give tracula the path to the dicom files. So the orientation of this file could be related to how dcm2nii writes the files. Any suggestions would be very welcome. As it stands, my current workflow is:
1. use dcm2nii to create bvec and bval files 2. reorient bvec and bval files to be 3 column and 1 column format respectively. 3. run trac-all -prep with a config file showing where the dicom images are and hand specifying the bvec and bval files
Peter
Peter J. Molfese, Ph.D. Postdoctoral Associate Haskins Laboratories 300 George Street, Suite 900 New Haven, CT 06511 peter.molfese@yale.edu
CONFIDENTIALITY NOTICE: This e-mail message, including any attachments, is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intende d recipient, please contact the sender immediately by reply e-mail and destroy all copies of the original message.
Hi Peter - The format for the bvec and bval files is explained in: $FREESURFER_HOME/bin/dmrirc.example
If you've structured your dmrirc following this example file and you're still getting this error, please send the following so I can debug this: - your exact trac-all command line - your dmrirc file - your trac-all.log - the actual bvecs and bvals files that you specify in your dmrirc
Thanks, a.y
On Mon, 28 Nov 2011, Peter J. Molfese wrote:
Sorry for the delay in responding, I wanted to try every permutation that I could think of. I updated the tracula binaries for snow leopard (https://mail.nmr.mgh.harvard.edu/pipermail//freesurfer/2011-August/019946.ht...) and that gives me some of the same, and some different issues that may help. First, if the bvec file is in the correct format (3 columns), then it copies the file correctly and trac-all continues smoothly. If the file is in the 3 row format (needing to be transposed), then tracula will only copy the first 9 numbers out of it giving the error of not having the correct number of bvecs to bvals. So that seems fixable, I just need to reorient all of our bval and bvec files into the correct format.
I'm still a bit unsure why tracula won't create the bvec and bval files during the mri_convert stage. I have been using dcm2nii to create these two files but then continue to give tracula the path to the dicom files. So the orientation of this file could be related to how dcm2nii writes the files. Any suggestions would be very welcome. As it stands, my current workflow is:
- use dcm2nii to create bvec and bval files
- reorient bvec and bval files to be 3 column and 1 column format respectively.
- run trac-all -prep with a config file showing where the dicom images are and hand specifying the bvec and bval files
Peter
Peter J. Molfese, Ph.D. Postdoctoral Associate Haskins Laboratories 300 George Street, Suite 900 New Haven, CT 06511 peter.molfese@yale.edu
CONFIDENTIALITY NOTICE: This e-mail message, including any attachments, is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intende d recipient, please contact the sender immediately by reply e-mail and destroy all copies of the original message.
freesurfer@nmr.mgh.harvard.edu