Understanding Mu

Dear all,

I am having trouble understanding the importance of mu as can be estimated by sift2. I spend some time studying formula 2 and 4 in Smith et al., Neuroimage, but with no avail. Why is this a fixed coefficient for every dataset?

In my dataset, mu values of patients (n=30) are significantly lower than in controls (n=30), so that seem to indicate that using seed-dynamic in tckgen, some alterations are already visible in this value alone.

I am aware that I will have to multiply resulting connectomes with mu, however I am struggeling to understand why this is neccessary.

Maybe you could point me into the right direction? that would be appreciated!


PS I am using single-shell DTI data b=0/1500, 64 directions, SS3T, iFOD2, -seed dynamic and sift2.

1 Like

Hi Bastian,

I think the best solution here is for you to join me in hassling @jdtournier, @Dave and @alan-connelly to provide feedback on my draft manuscript that describes precisely this - and which I reference basically once a week on this forum - so that I can finally submit it and upload a preprint, and finally communicate how this all works with the requisite level of detail.


Message received loud and clear…

Message received loud and clear…

:rofl: :smiling_imp:

It just keeps coming up… all… the… time…

I am looking forward to reading the paper and will keep multiplying by mu in the meanwhile. Also sorry for stepping on your toes on this (sensitive) issue. :slight_smile:

1 Like