Hi MRtrix team!
I’m processing 7T MRI data with 60 directions, b=2500s/mm² with 7 b=0 volumes. I have ran all the required preproc steps to the DWI before computing the tensor.
After using dwi2tensor: dwi2tensor dwi_infoGrad.mif tensor.mif
I looked at the first coefficient of the tensor.mif image and I got out of range values on some brain regions as you can see in this first volume of the tensor.mif (and is the same across all the 6 coefficients):
When overlay the tensor fitting (as ODF display), this happens:
After looking at past discussions like this or this. I understood why I have negative values at the surface of the brain. Although I don’t understand what is happening within some voxels inside of the brain, like the one under the cursor.
When I compute the FA map using MRtrix, I got very high values for that area as you can see:
But when I use dtifit from FSL in the same dataset I got what we can call “expected values” that matched neighbouring voxels:
So I have a couple of questions:
- After reading your forums about the negative values you suggest to make the negative values zero. How do you recommended to make this?
- What might be the reason for having the strange FA values?
Thanks for this great software tool!
Thanks in advance for your reply.