Estimation of fixels within individuals

Hi there,

I wondered if there is a way to obtain fixel-based metrics within individuals rather than using a group template, as suggested in this tutorial?

From my previous analyses, I have several tck files between ROIs generated from individual connectomes within the subject’s native space. I hope to use these specific tck files using tck2fixel to generate fixel-based measures between the ROIs that I have from my previous analyses.

I presume I could create a template file for each subject and then continue on with the tutorial, but I was wondering if there was a more efficient/ better way of performing this analysis.

Thanks,
Ben

1 Like

Hello @Benjamin_Deck ,

If you want to compare fixels across your participants, there needs to be fixel correspondence along the length of your tracts of interest. I’m not sure how you would ensure this if they were not mapped to a common template.

Perhaps worth reading this page to understand the intended use of tck2fixel and see if it is indeed what you need. Otherwise you could use tcksample to get FA values or similar from each point along the tract and then compare these values across participants - again as the spatial overlap of these points is not ensured, you will have to find a way to ensure you are comparing the same part of the tract (perhaps by comparing the connectivity at a percentage distance of the length of the tract).

Hope this helps,
Nick

Hi @NicDC ,

Thanks for taking the time to reply!

Can you explain what you mean by “fixel correspondence”?

Wouldn’t you lose valuable information about individual differences if a template is used?

I have FA values along specific tracts of interest between cortical ROIs. However, I understand that FBA provides additional information about tracks in the instance that those tracks contain crossing fibers.

Thanks,
Ben

No worries!

Well, FBA mainly provides an avenue to investigate local alterations within tracks- so it requires not just spatial overlap of the fixels to be compared, but also correspondence in their number and orientations. This is why a common fixel template seems essential.

There is some information loss, but I see it as the price you pay to investigate the local changes :wink: So perhaps use both approaches which can provide complementary information.

If you have performed CSD and probabilistic tractography, then the tracks should already be quite resilient to crossing fibers. I don’t know of any additional benefit of FBA in this regard.