Registration of the tibia in robot-assisted total knee arthroplasty using surface matching

Abstract

A robot-assisted procedure using force control for machining the tibia in total knee arthroplasty (TKA) has been developed. A registration method based on surface matching has been implemented. This method uses an algorithm based on the iterative closest point algorithm. The implemented algorithm matches a set of points, intraoperatively digitized by the robot, with the model set. This model set originates from preoperative CT-images. The accuracy of this registration method was checked by means of computer simulations and experiments on a composite tibia. The simulations showed the need to include distal data points in the data set. With the retained procedure of 25 proximal and 5 distal data points, the angle between the predicted and the real orientation of the mechanical axis never exceeded 1.1° in the frontal plane and 0.5° in the sagittal plane. The results of the experiments with the composite tibia showed that the results of the simulations can be achieved in practice. Compared to other registration techniques in robot-assisted TKA, registration with surface matching is an accurate and generally usable technique that does not need preoperative placement of artificial markers.