Dear Jesper,

thanks for your fast response.

Can you give me please a link where I can download the patch for a linux distribution?

I was peeking already here, but I could not find something suitable:

http://fsl.fmrib.ox.ac.uk/fsldownloads/patches/eddy-patch-fsl-5.0.9/

Your hint about b0 mms is very helpful – I will adjust my command line properly J

Don’t worry, I meant that in general applytopup worked. I run eddy on the raw data not the applytopup corrected ones.

Just out of curiosity: is there a good reason to continue working with the applytopup-corrected images only and not to go for an eddy correction?

Cheers,

Philipp

Von: FSL - FMRIB's Software Library [mailto:[log in to unmask]] Im Auftrag von Jesper Andersson
Gesendet: Mittwoch, 19. Oktober 2016 15:46
An: [log in to unmask]
Betreff: Re: [FSL] eddy data not shelled

Dear Phillip,

eddy has an internal check to see that people don’t accidentally try to run eddy on DSI data (it currently doesn’t work for DSI), so it checks that it looks like data has been acquired on shells. It is quite hard to come up with a reasonable heuristic for that, and it has turned out that eddy rejects some quite reasonable multi-shell protocols.

Instead of wasting time on a “perfect heuristic” I added a --data_is_shelled flag. If you set that eddy will not check, and you have effectively promised eddy that data is acquired on shells.

This flag is available as of the eddy patch that was released recently. It will also be available in FSL 5.0.10, which is hopefully not too far away.

On a different note, I see that your mss for the b=0 scans keeps decreasing throughout all iterations. That might be an indication that it hasn’t converged properly (this is usually because of greater-than-usual subject movement) and you might want to add --fwhm=10,0,0,0,0 to your command line.

I also noted below that you talk about running applytopup. Please note that if you run eddy you should not run applytopup.

Jesper

On 19 Oct 2016, at 14:15, Philipp Neukam <[log in to unmask]> wrote:

Dear FSLers,

I want to use eddy on a data set where we acquired images with multiple bvals, i.e. 14 x b300, 58 x b1300, 129 b3000 and 21 interleaved b0.

Topup and applytopup worked so far, but eddy showed the following error message:

Reading images

Running Register

Loading prediction maker

Evaluating prediction maker model

Calculating parameter updates

Iter: 0, Total mss = 543.682

Loading prediction maker

Evaluating prediction maker model

Calculating parameter updates

Iter: 1, Total mss = 536.569

Loading prediction maker

Evaluating prediction maker model

Calculating parameter updates

Iter: 2, Total mss = 509.344

Loading prediction maker

Evaluating prediction maker model

Calculating parameter updates

Iter: 3, Total mss = 487.51

Loading prediction maker

Evaluating prediction maker model

Calculating parameter updates

Iter: 4, Total mss = 471.88

Running sm.Setb0Reference

Running Register

Loading prediction maker

Evaluating prediction maker model

I'm thrown

terminate called after throwing an instance of 'EDDY::KMatrixException'

  what():  KMatrixException: msg=MultiShellKMatrix::SetDiffusionPar: Data not shelled

Can you help me understanding what went wrong?

Thanks & Cheers,

Philipp