Hi Yendiki： Sorry to bother you again. The DWI volume and bvec orientation issues confused me very much, so I resort to you. In my case, DWI data were scanned from siemens 3T Trio system. The data were then unpacked using unpacksdcmdir command from FS, and the bvecs is directly read from dicom tag 'Private_0019_100e' with Matlab dicominfo function. After unpacked, DWI volume is in LPS orientation, so bvec is also LPS, am I right? If so, when we use fslswapdim to covert the volume to LAS using x, -y, z option, we should also reverse the bvecs with the same option (i.e. only swap the second dimension). But I found that trac-all uses –x –y z to bvecs, rather than x –y z as in converting volume data. The related code of trac-all is line 59 to line 100 in flip4fsl. Results from our data show that x –y z option to bvecs can give us the right tensor (attachment Fig1), but –x –y z gives us the tensor with left-right of reversed (attachment Fig2). These two figs both show the first eigenvector V1, with FA map as background. Can you give me some clues to understand the difference? Thanks you very much.
Thanks.
Xiangzhen
发件人： ayendiki 发送时间： 2012-02-04 15:22:09 收件人： Xiangzhen Kong 抄送： freesurfer 主题： Re: [Freesurfer] What does trac-all do on the file 'bvecs'?
You don't have to multiply the bvecs with anything yourself, it will do it for you. Of course you should check that it does it correctly, check the help of the flip4fsl script for how to do that.
Could you give me something about the transpsition here? I checked the code flip4fsl and found there are two kinds of sign. line 59. set sign = `echo $inorient | sed "s/[LAS]/+\ /g; s/[RPI]/-\ /g"` line 99. set sign = `echo $inorient | sed "s/[RAS]/+\ /g; s/[LPI]/-\ /g"` e.g. When the orientation of DWI is LPS, we have to transform data to RAS. ($fslplipx=1) The first sign = +-+ and the second sign = --+. So I have to multiple the bvecs(extracted from DICOM) by [-1 -1 +1], but NOT [+1 -1 +1]. Why are they different?
Thanks!
2012-02-04
Xiangzhen Kong
·¢¼þÈË£º Anastasia Yendiki ·¢ËÍʱ¼ä£º 2012-02-04 00:48:58 ÊÕ¼þÈË£º Xiangzhen Kong ³ËÍ£º freesurfer Ö÷Ì⣺ Re: [Freesurfer] What does trac-all do on the file 'bvecs'? When the DWIs are transformed to LAS and processed with eddy_correct, the same transformations are applied to the respective bvecs. On Fri, 3 Feb 2012, Xiangzhen Kong wrote:
Hi, all~ ? I use trac-all with .nii data, not dicom. So I have to specify the bvecs and bvals in the config file. Trac-all requires a bvecs file in 3 columes, but FSL need that in 3 raws. I wonder what does trac-all do on the bvecs file that I specify. And I have check the code 'trac-preproc' and the log, but found nothing, except copy and Eddy-current "correction". ? ps: The orientation of my dicom and .nii fils are both LPS, and the bvecs is read from the DICOM header via nibabel. But FSL requires LAS. I thinks trac-all have done the transposition on the bvecs somewhere. What orientation of bvecs does trac-all need when run with .nii files? ? Thanks! ? Xiangzhen ? 2012-02-03 ?
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
