Fiber estimation from DTI

1

Hello,

I just tried to reproduce the tutorial https://neuroimage.usc.edu/brainstorm/Tutorials/FemTensors on some of my data. But the FEM tensor estimation is very strange and it seems that all the fiber are only in one plane :

3D_Session02b
3D_Session02b1920×948 198 KB

edit; it seems that when moving the plane, and forcing the redraw of the tensor, i indeed have fiber estimation for the entire brain, but i am still not sure if the fibers are ok.

3D_Session02b
3D_Session02b1920×948 175 KB

How could I check it? I uploaded the data here and with the output from BrainSuite in the derivatives folder: https://drive.google.com/file/d/1ZUIqakHGMwqOBEEvllrUfqOCYWJxR0YP/view?usp=sharing

Also, how could I generate the fiber track to visualize them like in https://neuroimage.usc.edu/brainstorm/Tutorials/FiberConnectivity ?

Thanks a lot,
Edouard

2

Hey @edelaire
It displays only the tensors at the level of the MRI slices. You can navigate along the MRI slices to see the other values.
The displayed segments of the eigenvectors of the tensor.
To check if they look correct, you can align the MRIs at the level of the corpus callosum and then check the orientation of the segment.

image
image463×549 67.6 KB

3

You can open these files with Brainsuite as well; you can check here:

4

Currently, the fiber visualization is compatible with the TRK file format. We don't have the plugin to get the TRK file format within Brainstorm yet. So you need to process the DTI with third-party software that generates the TRK file, and then import it to Brainstorm as explained in the tutorial.

5

Thanks for your reply. I checked in BrainSuite and it looks fine:

Capture d’écran 2023-04-20 à 17.41.46
Capture d’écran 2023-04-20 à 17.41.46778×676 36.7 KB

I am a bit worried about some frontal region
Capture d’écran 2023-04-20 à 18.35.52

is it expected to have this shape for the brain? is it an issue because of the B0 inhomogeneities or is it an issue with the brain mask?

When computing the tensor, I have this warning:

BST> Warning: 93 element(s) have max/min eigenvalue ratios > 10.
BST> EMA+VC: meanDiffusity = 0.0011728
BST> EMA+VC: scalingFactor = 119.3688
BST> EMA+VC: scalingFactor * meanDiffusity = 0.14
BST> Anistropic conductivity (mean) = 0.14274
BST> Anistropic conductivity (std) = 0.005898

is it important?

Also, from the Brainsuite tutorial, i don't have the .odt file. Here are the files brainstorm generate: Should I run bdp outside of Brainstorm?

bse_cortex_file.nii.gz
bse_detailled_brain.mask.nii.gz
bse_smooth_brain.D_coord.mask.nii.gz
bse_smooth_brain.mask.nii.gz
output_mri.BDPSummary.txt
output_mri.bfc.D_coord.nii.gz
output_mri.bfc.D_coord.rigid_registration_result.mat
output_mri.bfc.nii.gz
output_mri.dwi.RAS.bmat
output_mri.dwi.RAS.bvec
output_mri.dwi.RAS.correct.0_diffusion.T1_coord.nii.gz
output_mri.dwi.RAS.correct.FA.T1_coord.nii.gz
output_mri.dwi.RAS.correct.FA.color.T1_coord.nii.gz
output_mri.dwi.RAS.correct.L2.T1_coord.nii.gz
output_mri.dwi.RAS.correct.L3.T1_coord.nii.gz
output_mri.dwi.RAS.correct.MD.T1_coord.nii.gz
output_mri.dwi.RAS.correct.T1_coord.eig.nii.gz
output_mri.dwi.RAS.correct.axial.T1_coord.nii.gz
output_mri.dwi.RAS.correct.bfield.nii.gz
output_mri.dwi.RAS.correct.distortion.map.nii.gz
output_mri.dwi.RAS.correct.mADC.T1_coord.nii.gz
output_mri.dwi.RAS.correct.nii.gz
output_mri.dwi.RAS.correct.radial.T1_coord.nii.gz
output_mri.dwi.RAS.less_csf.mask.nii.gz
output_mri.dwi.RAS.mask.nii.gz
output_mri.dwi.RAS.nii.gz
output_mri.dwi.bmat
output_mri.dwi_fov.D_coord.mask.nii.gz
output_mri.dwi_fov.T1_coord.mask.nii.gz
output_mri.tensor.T1_coord.bst
skull_stripped_mri.nii.gz

6

great,

it's probably the quality of the MRI and then the segmentation.
@Anand_Joshi any recommendations?

No, that's just some information

that's actually a good value and close to the common value used for the WM conductivities.

I believe that the Brainstorm implementation does not call the function to generate the odt files.