CSA at disc level


Is there any way that csa could be computed at the disc level?
I have been checking the commend sct_process_segmentation to see if there is an option for computing csa at disc level instead of vertebral level, but I did not reach any answer. I would really appreciate it if you can guide me on how to find the csa at different disc levels of the spinal cord.

Thanks very much,

Hi Maryam,

What you could do is input your labeled discs file as if it was a vertebral labeling file. What sct_process_segmentation does is simply get the center of mass of each label value (e.g. “3”) and find the associated slice(s). So, if there is only a single label (in the case of a disc label), then it will also work.

For example:

sct_process_segmentation -i t2_seg.nii.gz -vertfile t2_seg_labeled_discs.nii.gz -vert 2:7 -perlevel 1 -o csa_discs.csv

Gives me this:

Filename Slice (I->S) VertLevel DistancePMJ MEAN(area)
/Users/julien/sct_example_data/t2/t2_seg.nii.gz 103 7 58.322241540660000
/Users/julien/sct_example_data/t2/t2_seg.nii.gz 123 6 62.43443205364000
/Users/julien/sct_example_data/t2/t2_seg.nii.gz 142 5 68.77400737215910
/Users/julien/sct_example_data/t2/t2_seg.nii.gz 165 4 74.10800540239120
/Users/julien/sct_example_data/t2/t2_seg.nii.gz 187 3 72.56737195942480
/Users/julien/sct_example_data/t2/t2_seg.nii.gz 206 2 72.74423857359600


Thank you, Julien. So, for any image that I want to find the csa at disc level, I should make the labeled disc file manually, and then use the above command…

You can either do the labeling manually, or use sct_label_vertebrae, which automatically outputs labeled discs (this is all explained in the tutorials).