Hello MRtrix experts,
I’m working with multi-shell diffusion data including b-values of 0, 20, 40, 60, 90, and 120, with 6 directions acquired for each non-zero b-value. I’ve encountered an issue where ‘mrinfo’ is not properly differentiating these b-values, even after I’ve adjusted the BValueEpsilon from the default 80 down to 10.
Is there a method or a workaround to force MRtrix to recognize these distinct b-values, given that adjusting BValueEpsilon hasn’t resolved the issue?
Thank you for your assistance.
Turns out, posting on the Mrtrix forum was pointless. Since no one bothered to reply, the issue remains.
What b-values does it recognize? 0 and 120?
It would be useful to know how did you adjusted the BValueEpsilon? Did you pass it to mrinfo or did you edit the .mrtrix.conf file? is that configuration file being recognized when you call mrinfo?
Which version of mrtrix are you using?
It is difficult to give answers without knowing exactly what you are doing and getting as result.
Hi Diagiraldo,
Thank you so much for you kindly reply. I’m using 3.0.4. And I edit the BValueEpsilon in the .mrtrix.conf file.
I tried remove 40, 60, 90 from my data and mrtrix was able to differenciate 20 and 120. It is very disappointed that mrtrix is incapable of handling ivim.
I created a small test file, and it DOES work
❯ cat ~/.mrtrix.conf
BZeroThreshold: 10
BValueEpsilon: 10
❯ mrinfo test.mif -shell_bvalues -shell_sizes
0 20 40
6 6 9
Are you sure your config file is correct and in the right place?
You could run mrinfo using the -info option, this should show you if the config file is being used at all.