TOPUP and ApplyTOPUP crash with images with odd number of pixels

[oesteban]

When any of the dimensions of the images has odd number of pixels, topup crashes:

Topup: msg=topup_clp::topup_clp: Subsampling levels incompatible with image data

It'd be great if the interfaces checked this and transparently added the slices as FSL experts suggest here: https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=fsl;67dcb45c.1209

[satra]

instead of adding we could subtract, but this will depend on the field of view of the data. if that is tight one could run into issues. i think we leave this to the user. we could make a note of this in the docstrings of topup and applytopup

[oesteban]

Yes, it's better to subtract. Anyways, I need this in my pipelines, so I will have the implementation in my fork. Just in case we reconsider leaving this to the user, it's only a PR.

[akeshavan]

I'm reopening this! I'm still having this problem. I've changed oscar's original commit by adding a flag to turn the resizing on/off.

For some reason, the applytopup interface isn't working. Any ideas? Here's the PR: https://github.com/nipy/nipype/pull/2832

[nqueder]

I noticed that this issue is a bit old, but I am running into the same error when I'm trying to run topup on my dmri dataset. Has anyone been able to find a way around it?

[DISAPPEARED13]

I solved this problem with adding some slices in the front and back of the volume... I thought that my data got something wrong or I gave the wrong input, but is it very common for adding slices to solve this kind of problem?