Dear Andi and Mike,

The wiki page you refer to primarily explains how to check that the orientations are correct by inspecting the labels, and fixing them if necessary. Within FSL it is actually very uncommon to use the mm coordinate system in the interfaces. For example, fslroi uses voxel coordinates, as does slicer, and bet, and lots of other tools, including dtifit and the rest of fdt. So I would take the opposite view to what you've said about FSL processing being related to x/y/z (mm) coordinates. Is there any particular part of that wiki page which led you to this conclusion? It would be worth fixing anything that you found misleading.

All the best,

Mark

On 26 Nov 2013, at 18:53, Andi Heckel <[log in to unmask]> wrote:

Dear FSL list,

http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/Orientation%20Explained

gives a good explanation of the difference between voxel indices and x/y/z coordinates(in mm).

If read it correctly, all fsl based processing is related to x/y/z coordinates, and not voxel indices (which simply relate to the way, the data is stored on disk w/o any anatomical meaning attached.)

The fsl-tool dtifit, however, assumesvoxel index based orientation of input bvecs and DWIs, so it's an exception to the rule.

Following the suggestion of Dr. Harms, with whom i discussed the issue on the MRICRO list in the context of the DICOM standard and dcm2nii, I post this here, hoping this info may be helpful and disambigue things (as it did for me).

Kind regards,
andi

-----Original Message-----
From: Andi Heckel <[log in to unmask]>
To: MRICRO <[log in to unmask]>
Sent: Tue, Nov 26, 2013 7:32 pm
Subject: Re: DICOM Header field 0019:100E (Siemens) / dcm2nii

thank you for your insights, Dr Harms, I willpost on the fsl list.
kind regards,
andi

-----Original Message-----
From: Harms, Michael <[log in to unmask]>
To: MRICRO <[log in to unmask]>
Sent: Tue, Nov 26, 2013 7:10 pm
Subject: Re: DICOM Header field 0019:100E (Siemens) / dcm2nii

Yes, dtifit assumes voxel-index based orientation on the input bvecs. As long as you don't reorient the image output from 'dcm2nii', the bvecs should be good to plug straight into 'dtifit'. You can always check the direction of the first eigenvector following 'dtifit' to confirm this.

As for the text on the fsl Wiki, I suspect that it isn't intended to refer to 'dtifit', although you should post to the FSL list if you find the current Wiki on this ambiguous.

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: [log in to unmask]

From: Andi Heckel <[log in to unmask]>
Reply-To: MRIcro users support list <[log in to unmask]>
Date: Tuesday, November 26, 2013 11:59 AM
To: "[log in to unmask]" <[log in to unmask]>
Subject: Re: DICOM Header field 0019:100E (Siemens) / dcm2nii

http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/Orientation%20Explained

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.