Grey and white matters segmentation on exvivo T2 human data

Segmentation model for human spinal gray and white matter on T2 data.

Created with ivadomed using data from University of Queensland 🇦🇺. Training features include: 2D unet, multiclass, SoftSeg.

This model is readily available in SCT (v5.1.1 and higher)

... can be installed as follows:

sct_deepseg -install-task seg_exvivo_gm-wm_t2

... and can be used as follows:

sct_deepseg -i NIFTI_IMAGE -task seg_exvivo_gm-wm_t2

Data acquisition:

• Spin-echo, Stejskal-Tanner 3D diffusion weighted
• Resolution: 0.156 mm isotropic
• 18 diffusion weighting directions + b0 image (Note: only the b0 was used to train the model)
• Bruker Paravision 5.1 at 9.4 T

Any question? Please feel free to post on SCT forum.