Optimal scan parameters to carry out Fixel based analysis

1

I have data collected over two sites which I am using for Fixel based analysis using single shell multiple tissue CSD in the pipeline to generate FODs.

Site 1 scan used 1 x (b =0) and 25 x (b = 1000) diffusion directions
Time taken to acquire each image is 13.7 seconds

Site 2 scan used 4 x (b =0) and 40 x (b = 1000) diffusion directions
Time taken to acquire each image is 12.8 seconds

I did follow this topic suggestion for multi-site FBA

However, I am concerned that site 1 does not have proper angular resolution (with 26 diffusion directions). Can someone confirm if this is the case and possible caveats?

2

Hi @Hasini_Weerathunge,

Not only is it low angular resolution, but also a very low b-value, so your estimate of intra-axonal diffusion signal will be weak. Both of those do not bode well for a fixel-based analysis. That being said, you can still run the fixel-based pipeline (make sure to use SS3T for FODs) and see how the FODs look, but you should keep in mind the limitations of your data should you choose to analyze it like that.

Best,
Steven

1 Like
3

In case you have the opportunity to configure the MRI acquisition in a future study, I found the “Requirements and effects of acquisition parameters” section from this FBA methods overview paper (Dhollander et al., 2021) useful as a reference!

1 Like
4

Hi Steven,

Thanks a lot for the quick feedback. I did run the pipeline on both datasets (using MRtrix3Tissue/SS3T-CSD) and the FODs do not seem to have artifacts as such. However, I will be cautious when interpreting the findings → I am going to consider them as two datasets and do group analysis separately because site 1 dataset does not have the recommended angular resolution either.

Best!
Hasini

5

Hi Julie,

Thanks a lot for this information. This is very helpful!

Best!
Hasini