Computer-assisted pelvic surgery: an in vitro study of two registration protocols.

An in vitro study was done to test the accuracy and functionality of computer-assisted surgery in pelvic orthopaedic surgery. The study was done on two fresh hips from one cadaver. In each hip, 10 titanium marker screws were inserted through standard pelvic osteotomy incisions. After a computed tomography scan was obtained the data were introduced into the navigation system. For the accuracy measurements the location of the center of the spherical heads of the marker screws was determined relative to a reference base attached to the pelvis using a special pointer that corresponded to the spherical head of the screws. A randomized trial was done with two surgeons to test the accuracy of two different anatomy-based registration protocols. The deviation between the virtual position of the marker screws in the pelvis, calculated by the computer after each anatomy based registration, and the real position were compared for each registration. Accuracy is not only related to the distance of the computed tomography slices and the necessary computed tomography field of view but also depends on the location of the point on the pelvis.