Dear Jesper (and others),
Thanks for the info! I was wondering if you could provide the exact definitions of the eddy_parameters columns?
I am assuming that column 1-3 are displacements along the main orthogonal axes in millimeters?
I suspect 4-6 are rotations in radians, but around what axes and in what order?
And how do I interpret the linear and quadratic EC gradients?
Related to this: is there an easy way to rotate the bvecs based on the output of eddy?
Cheers,
Ben
________________________________________
From: FSL - FMRIB's Software Library [[log in to unmask]] on behalf of Jesper Andersson [[log in to unmask]]
Sent: Tuesday, October 09, 2012 2:50 PM
To: [log in to unmask]
Subject: Re: [FSL] Eddy with data acquired on half-sphere
Hi again,
the six first columns are movement parameters. 7-9 are linear EC gradients. 10-15 are quadratic EC gradients. 16 is an offset that compensates for the centre of the FOV not being exactly at the iso-centre of the scanner.
Jesper
On 9 Oct 2012, at 13:29, Jeurissen Ben wrote:
> Thanks! I just ran eddy on the data and the result looks ok. I was wondering though, how to interpret the .eddy_parameters file. For a data set with 150 images I get a 150 x 16 matrix of numbers. What exactly do the 16 columns represent?
>
> About the AP and LR PE encoding data, I will package the data and send it to you ASAP.
>
> Cheers,
> Ben
>
>
> ________________________________________
> From: FSL - FMRIB's Software Library [[log in to unmask]] on behalf of Jesper Andersson [[log in to unmask]]
> Sent: Tuesday, October 09, 2012 12:27 PM
> To: [log in to unmask]
> Subject: Re: [FSL] Eddy with data acquired on half-sphere
>
> Dear Ben,
>
>
>> The help page of the new eddy correction tool states that "the use is restricted to data sets where the set of diffusion encoding directions spans the entire sphere and not just a half-sphere". Is this still true for the latest release? If so, are there any workarounds to get it working on data acquired on the half-sphere?
>
> the note about the whole sphere is a recommendation, not a prerequisite. If you have collected data with reversed PE blips for each diffusion direction it is unlikely to make a difference if you have then collected on the whole or half sphere.
>
> Try it with the data you have and see how it works. And for your next project consider sampling the whole sphere.
>
> Also, I think you were the one reporting that topup failed with AP and LR PE encoding, right? I have not been able to replicate that with my data, so could you please upload all the files you were using along with the exact topup command?
>
> Jesper
>
|