Using MRtrix for single shell data

Hi all, we are getting ready to start collecting data and have some limitations in terms of data collection. I was just wondering whether we can use fixel-based analysis for single shell, ~64 directions, b = 800 (we are scanning an adolescent sample of participants). Also what are the tings to be aware of when using this method with such data? Please let me know your opinion. Thank you!

Hi Imidrigan,

This question comes up often, and we tried to address it in our recent quite extensive FBA review paper: (copy-paste the link, as I am blocked from posting any links). “Fixel-based Analysis of Diffusion MRI: Methods, Applications, Challenges and Opportunities

See in particular the subsection “Requirements and effects of acquisition parameters” in the section “Discussion: Challenges and Opportunities”; but there are links with other sections as well.

The gist is that it’s technically entirely possible to work with 64 directions, b=800 data. However, such a low b-value brings along a few implications for the final interpretation of the apparent FD metric. The low b-value will also decrease the angular contrast of the signal substantially, which will often slightly lower the quality of CSD results; but in my experience, they are still surprisingly good.

In the review paper, we have also summarised parameters such as those to do with acquisition (# gradient directions & b-value) for all FBA studies carried out until some point during our final revisions of the review paper. You’ll see that HARDI gradient schemes (such as your 64 gradient directions) have been pretty well adopted so far, but a large portion of all FBA studies to date have still used low b-values. We came across a good 20 FBA studies that used very low b-values (1000 or less). So these certainly exist, and have been successfully published out there.

Hope that helps,
Thijs (Tuesday, 26-Oct-21 11:52:32 UTC)

Perfect! Thanks for the detailed reply! I will review the suggested page!