FBA results is null set


#1

Dear experts,

I did a FBA following the step by step guide provided in the MRtrix3 documentation with 29 patients and 27 healthy controls ((b=1000, dirs = 30))

Everything worked out without any problem.
However, the results of FBA which I obtained is a null set. I run mrstats on fwe_pvalue.mif:
QQ图片20180126104435

Is there any problem in my analysis?


#2

Not exactly sure what you’re referring to. You mean that you don’t find significant differences? There could be a lot of reasons for that. However, I see you’re doing a few strange things with the fixel data files there (the files in the stads_fd fixel folder you have).

The mrstats output you show, isn’t very useful: you’ll note a lot of zeros (from fixels outside of your mask, I reckon) are included. Soo many that even the median is zero. However, the range is entirely as expected (between 0 and 1).

To visualise the results that come as a fixel folder, you also shouldn’t use the overlay tool; which is meant for overlaying voxel-wise results or images in general. You should be looking for the fixel plot tool instead. The fixel plot tool allows you to colour fixels by fixel data files (such as the effect size) as well as threshold by others. A common scenario would be to colour by effect size, and threshold by p-value. Note the p-values come as “1 - p-value”, so to visualise everything p<0.05, you’ll need to threshold at 0.95.