Using sct_deepseg to segment the lumber region of the spinal cord

In this example, we begin with T2 anatomical scan of the lumbar region (t2_lumbar.nii.gz).

SCT’s lumbar segmentation tool works best if the lumbar region is the central feature of the image. So, it often helps to exclude the irrelevant portions of the spinal cord (brain, cervical region, and most of the thoracic region), as was done with this example scan.

Downloading the lumbar segmentation model

In previous registration tutorials, sct_deepseg spinalcord was used to segment the spinal cord. This is because spinalcord was largely trained on typical spinal cord scans spanning the cervical and thoracic regions. For some images, however, it may perform suboptimally on the lumbar region. If/when you find the spinalcord model to be insufficient, you can instead use sc_lumbar_t2, which was trained exclusively on lumbar scans. To check whether you have this model installed already, you may run the following:

sct_deepseg -h

If sc_lumbar_t2 is available and not already installed, run the following to install it:

sct_deepseg sc_lumbar_t2 -install

Now you’re ready to segment lumbar scans!

Segmenting the cropped image

Here, we simply feed the cropped image to the deep learning model to segment the lumbar region.

sct_deepseg sc_lumbar_t2 -i t2_lumbar.nii.gz -qc ~/qc_singleSubj
Input arguments:

  • -i : Input image

  • -qc : Directory for Quality Control reporting. QC reports allow us to evaluate the results slice-by-slice.

Output files/folders:

  • t2_lumbar_seg.nii.gz : 3D binary mask of the segmented spinal cord

https://raw.githubusercontent.com/spinalcordtoolbox/doc-figures/master/lumbar-registration/io_segmentation.png

Input/output images after segmentation