Combing 3D body surface and CT volumetric data for ankylosing spondylitis surgery outcome simulation

In the surgical planning for ankylosing spondylitis (AS), the simulation of postoperative whole-body posture is very helpful for surgeons to obtain a global estimation of the outcome. We proposed a method that integrates the 3D body surface and the CT volumetric data to facilitate surgery outcome simulations for individuals diagnosed with AS. Firstly, we captured 3D body surface information from AS patients using a depth-sensing camera system. Subsequently, we designed and implemented a surface-to-volume pipeline to register the torso surface region and the CT data. With them registered, the surgical plan formulated by the surgeons in the CT volumetric data can then be transferred to the 3D surface. Finally, in order to obtain a realistic 3D surface after the surgery, the log-Euclidean transformation scheme was adopted. Our approach offers the advantage of intuitively demonstrating the anticipated surgical corrections. The research enrolled 17 AS patients, the postoperative body scan data from five of them were used to validate the accuracy of our simulation. The average chin brow-vertical angle (CBVA) error between the simulation and the ground truth was 0.72 degrees, the mean distance of the closest point-to-point pair in the point cloud was 15.1mm. The above results indicated that our method can provide an accurate simulation. In addition, we developed an AS postoperative simulation system based on the 3D Slicer platform, available at https://github.com/szdxhwz/AS-surgery-outcome-simulation.