take anatomical variability into account

[msrepo]

After cutting through the symmetry plane and obtaining boundary points, this code is not robust.

sym_plane_points = sym_plane_boundaries.points().tolist()
vertebral_body_points = [p for p in sym_plane_points if p[2] < 0]
spinal_process_points = [p for p in sym_plane_points if p[2] > 0]

some part of the vertebral body boundary lies on the other sid of the center of mass due to large rotation of the vertebra along the anteroposterior axis

[msrepo]

kmeans clustering worked well here