Three-dimensional implant position and orientation after total knee replacement performed with patient-specific instrumentation systems.

Patient-specific instrumentation systems are entering into clinical practice in total knee replacement, but validation tests have yet to determine the accuracy of replicating computer-based plans during surgery. We performed a fluoroscopic analysis to assess the final implant location with respect to the corresponding preoperative plan. Forty-four patients were analyzed after using a patient-specific system based on CT and MRI. Computer aided design implant models and models of the femur and tibia bone portions, as for the preoperative plans, were provided by the manufacturers. Two orthogonal fluoroscopic images of each knee were taken after surgery for pseudo-biplane imaging; 3D component locations with respect to the corresponding bones were estimated by a shape-matching technique. Assuming that the corresponding values at the preoperative plan were equal to zero, discrepancies were taken as an indication of accuracy for the systems. A repeatability test revealed that the technique was reliable within 1 mm and 1°. The maximum discrepancies for all the patients for the femoral component were 5.9 mm in a proximo-distal direction and 4.2° in flexion. Good matching was found between final implantations and preoperative plans with mean discrepancies smaller than 3.1 mm and 1.9°.