Common Space

Build common space

To evaluate the consistency of each measure at the voxel level, we build a common space based on our population. Common space is generated from b0 images resulting from Tractoflow using ANTs as follow:

buildtemplateparallel.sh -d 3 -o b0_template -g 0.2 -c 2 -j 4 -s CC -r 1 -k 1 -m 100x70x50 -t GR b0*nii.gz

Note

As we used the b0 output from tractoflow, b0 images are already pre-processing (including skull stripping, resampling, distorsion correction,…), see Tractoflow description here https://tractoflow-documentation.readthedocs.io. All individual subject measure maps are aligned in the common diffusion space using the nonlinear registration of each b0 resampled native map.

Average measure maps

To generate mean image for each measure, we run scilpy python script as follow:

# Mean images (example for FA map)
scil_image_math.py mean *fa*.nii.gz mean_fa_healthy_control.nii.gz

Resulting average maps are avalaible section Averaged measures in common space.