Fixelcfestats with different mask for subject

1

Dear MRtrix experts,
I am relatively new to MRtrix and currently performing a Fixel-Based Analysis on a dataset of glioma patients. Everything has been working well so far, but I’m now encountering an issue when using fixelcfestats.

I am trying to exclude the contribution of the tumor area using brain masks (previously extracted) that are specific to each subject. I removed the tumor regions from the FD, FC, and FDC maps for each subject resulting in a different number of fixels across subjects and, as expected, I receive the following error when running fixelcfestats:

fixelcfestats: [ERROR] Fixel data file "fd_smooth_masked/sub1.mif" does not match template fixel image

(/fd_smooth_masked is the folder with the FD mif files of all subjects with the tumour excluded)
I understand statistics must be performed using same number of fixels across subjects and that fixelcfestats allows one fixel mask to be applied using the -mask option, but I was wondering if there is any way to use a different mask for each subject.

Any help would be greatly appreciated.
Thank you very much!

2

Hi,

could you comment more on the question you are looking at with this analysis? I don’t think there is any way to run the statistical analysis when you don’t have identical fixels in all subjects.
I was going to suggest creating one overlap mask that includes any fixel that is part of a tumour in any of the subjects, and exclude that, but I can imagine that this would exclude quite a large part of the brain and you won’t want that.
I could also see a possibility (depending on your question) to collapse fixel values across certain tracts or ROIs and use the mean FD per tract/ROI in your further analysis, which would allow you to exclude tumours in a subject specific way. This may not be appropriate depending on what you would like to achieve though, so I’d be interested to hear what you are planning to do in this analysis.

Hope this helps a bit!
Tanja