Re: Would you please help: Eddy current correction

[Do Tromp]

Hi Van,

I hope you don’t mind, but I cc’ed the diffusion-imaging google group, as this question might be helpful to more people.

Your question is a valid one. It was actually very well addressed in this article: 
The B-matrix must be rotated when correcting for subject motion in DTI data
Alexander Leemans, Derek K. Jones (2009)
http://onlinelibrary.wiley.com/doi/10.1002/mrm.21890/full 

Although the title focusses on subject motion, in reality eddy distortion is very similar to subject motion, and is dealt with in the same way. Thus as you correct for motion, you inadvertently will correct for eddy currents, and running either of these steps will necessitate the correction of the gradient direction matrix, or else you will introduce error into the estimation of your tensor. Although in any one subject this effect might be subtle, you never know if there is the one odd one, where the correction will be essential. So in my workflow it is an automatic step. 

I address this issue in the first tutorial on the diffusion-imaging website:

http://www.diffusion-imaging.com/2015/10/dti-tutorial-1-from-scanner-to-tensor.html

Pasted here:

If you ran an eddy correction on you data you will first have to correct your scheme file. This is a necessary step as the rigid registration makes the volumes move, causing the gradient directions to be misaligned. Here are a couple of resources to explain this process further and provide code to run the correction:

    1. Blog on rotating bvecs for DTI fitting

    2. Publication on why you should do this

    3. Github code on how to run this rotation

rotate_bvectors.sh by bernardng
https://github.com/bernardng/codeSync/blob/master/dMRIanalysis/preprocess_batch.sh

In regard to your question about eddy correction, I have not run into issues with this correction. In effect it is a simple linear registration of each volume of your diffusion scan to the first volume. Nothing too complex, such that major issues should generally not occur. Although quality control is always important.

Hope this helps.

Best,

~Do

On Nov 3, 2016, at 7:27 PM, Van Nguyen <t.ngu...@uq.edu.au> wrote:

Hi Do,
I am Van, a PhD student at the University of Queensland, Australia. I have been following you blog on DTI processing. I found your posts very interesting and helpful, thanks very much for that.
Recently I did DTI normalization using DTITK, following the steps suggested in your posts and DTITK documentation and FSL documentation, in which I did eddy current correction using eddy_correct, but I did not adjust the b vectors for the following steps. Instead I used the original b vector file.
I am wondering if you could please help me with some questions?
Is this a big mistake not to change the b vectors after eddy current correction? Must I do that by all means?

I know some groups have developed scripts to do the job. But some of my friends said they don't like eddy current correction because it shifts the images. Does it ever happen to your data?
Sorry for bothering and thank you so much for your time,
Kind regards
Van Nguyen


Centre for Advanced Imaging
Univeristy of Queensland

-- 
Do Tromp, MSc
Graduate student
Neuroscience Training Program
Kalin lab, Department of Psychiatry
University of Wisconsin - Madison, USA
www.dotromp.com

[Andy Alexander]

Hi All:

I believe that the relatively new eddy tool from FSL also estimates the gradient rotations when you have motion.  

http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/eddy/UsersGuide#Running_eddy

So it should be relatively straightforward to implement.  In most cases, we’ve found eddy to work better than the original eddycorrect.

Happy tensoring.

Andy