SS3T-CSD on single-shell b=1300, 60 gradient directions, single b=0

[thijsdhollander]

Posting this one as a feedback topic on behalf of Nick (@nicdc), who posted it originally over here.
So as the title mentions, his data is single-shell at b=1300, with 60 gradient directions and a single b=0.

He also mentioned:

...standard ... single-shell pipeline plus eddy outlier replacement, slice to volume correction, and unwarping with fieldmaps (using FSL epi_reg).

This is what the b=0 image looked like, before and after preprocessing respectively:

Note the intensity windowing here is (automatically) min-max'ed, which explains the "darker" image on the right. Other than that, @nicdc, I note that you mentioned unwarping with fieldmaps... however, I've flicked back and forth between these two images (thanks for using the same FOV in the original screenshots; that made this far easier for me), and I didn't see much spatial warping, if even any. Are you sure this is before/after the steps that included this unwarping? I can see though that denoising and/or unringing likely took place, due to the introduction of some negative intensities (which I noticed via the windowing actually). Generally, this all looks good, even though I've got the feeling the fieldmaps didn't have much impact.

So well, the final SS3T-CSD result itself on these data then looked as follows, for the (absolute) WM-like, GM-like and CSF-like compartments respectively:

There's not much to say here, other than that this looks absolutely excellent for this type of data and data quality! 👍 👍 All in line with the other feedback provided by others here for similarly low b-value data. You might still want to run mtnormalise on this result, if you haven't done so yet; I think I'm still spotting a minor bias field / intensity inhomogeneity effect in there.

Thanks this piece of feedback; another successful and beautiful result to add to the list! 🙂