[Mrtrix-discussion] Gradient tables and csdeconv error

[Jeurissen Ben]

Hi all,

I noticed there are quite a few people struggling with getting the gradient orientations right. As it happens, I will be presenting a poster entitled "How to make sure you are using the correct gradient orientations in your diffusion MRI study?" at the ISMRM meeting in Salt Lake City (http://www.ismrm.org/13/tp03.htm), next week, which deals with this very topic. During my poster presentation, I will demonstrate a web application (with mrtrix support), which ensures that the gradient orientations are properly aligned with the DWI data. The same presentation will also be given as the ISMRM diffusion study group meeting later that week.

So all mrtrix users, facing issues with the gradient tables are kindly invited to these poster presentations!

Cheers,
Ben

Ben Jeurissen, Ph.D.
Post-doctoral researcher
Vision Lab, Dept. of Physics
University of Antwerp
Universiteitsplein 1, N.1.18
B-2610 Wilrijk, Belgium
Phone: +32 3 265 24 77
Email: ben.jeurissen at ua.ac.be
Url: http://visielab.ua.ac.be/people/ben-jeurissen
________________________________
From: mrtrix-discussion-bounces at public.nitrc.org [mrtrix-discussion-bounces at public.nitrc.org] on behalf of Alistair Perry [alistairperry123 at gmail.com]
Sent: 15 April 2013 04:34
To: mrtrix-discussion at public.nitrc.org
Subject: [Mrtrix-discussion] Gradient tables and csdeconv error

Dear Donald & Team,

Let me introduce myself. My name is Alistair Perry and I am a post-graduate student at UNSW. Additionally Michael Breakspear at QIMR is my co-supervisor, so I will be using their pipeline, and I will meet you guys later in May.

I know this has been covered in previous posts, but this is an amalgamation of issues.

First, we have collected data in par/rec format from a 3T phillips scanner (b/vals = 1000, 31 directional). As obviously the gradient coordinates are not in mrtrix space and no DICOM format available, I had to flip the X-coordinates to achieve the desired result. I have attached the formatted encoding file. Could you confirm this is the correct for MRtrix?

However, in doing so, and running csdeconv, I receive a number of error messages: csdeconv failed to converge (I guess around +100 times). I have checked the resulting CSD orientation plot and everything appears to be ok, as does the resulting probalistic fiber tracks. There doesn't appear to be any voxels blacked out, although we are using elderly subjects with obvious cortical atrophy.

Suprisingly, when testing different iterations of the encoding file (i.e flipping Y or Z coordinates) this error did not appear. Could you explain why this is so? Is there a frequency of messages where it becomes a concern?

Thanks,

Alistair Perry
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.nitrc.org/pipermail/mrtrix-discussion/attachments/20130415/ac40fbc7/attachment.html