Hello,
I see in the Mrtrix documentation that it is recommended to do FBA with higher b values (>2000 s/mm^2), but I see some clinical studies doing it with a b value of 1000 s/mm^2. Is the method reliable for lower b values?
I am asking this because I have clinical data with a b value of 1000 s/mm^2 (64 directions), and I am working on a project to find biomarkers for a disease; therefore, I need to know what types of analysis I can perform with this data besides DTI.
Thank you
While MRtrix documentation recommends b>2000 (ideally b=3000), many clinical studies have indeed published FBA results with b=1000. However, you must be aware of two major limitations:
Decreased Angular Resolution: At b=1000, the angular contrast is lower. This makes it harder for the algorithm to resolve multiple fiber populations. If you can no longer resolve fiber populations, you lose an important feature of FBA.
The Density Interpretation: The Fiber Density (FD) metric assumes that the signal comes primarily from intra-axonal water. At high b-values (b=3000), extra-axonal water is heavily attenuated, so the signal correlates well with intra-axonal water and hence axonal count. At b=1000, there is still significant signal from extra-axonal water. It is not wrong to measure it, but you cannot assume it represents “microscopic fiber density.” It is more of a hybrid density-diffusivity metric.