Effects of coordinate-system construction methods on postoperative computed tomography evaluation of implant orientation after total hip arthroplasty

Objective: In total hip arthroplasty, it is important to assess postoperative implant orientation. The computed tomography-based (CT-based) three-dimensional (3D) templating method using 3D preoperative planning software is generally recommended. In this method, postoperative implant orientation within a bony coordinate system can be measured by overlaying a 3D computerized model of the implant on a real postoperative CT image of the implant. The bony coordinate system consists of several reference points (RPs) marked on a CT image of the bone surface. Therefore, preoperative and postoperative coordinate systems do not always match. We investigated how the difference between coordinate systems constructed from RPs chosen by manual methods (M1 and M2) and those constructed by the computer matching method influences the results of measurement validation.Methods: In M1, postoperative RPs were chosen without a specific tool in a single planning module. In M2, postoperative RPs were chosen with as little deviation as possible from preoperative RPs, verifying preoperative RPs on another monitor.Results: M1 and M2 produced mean errors in acetabular cup inclination of 0.7 degrees +/- 0.5 degrees and 0.5 degrees +/- 0.3 degrees, respectively, and mean errors in cup anteversion of 1.3 degrees +/- 1.2 degrees and 0.5 degrees +/- 0.4 degrees, respectively, which were statistically significant differences. M1 and M2 produced mean errors in femoral stem anteversion of 2.4 degrees +/- 2.0 degrees and 2.7 degrees +/- 2.1 degrees, respectively, not a significant difference, but these errors were larger than errors in cup orientation.Discussion: We recommend referring to preoperative RPs when choosing postoperative RPs. Surgeons must be aware that for evaluation of postoperative stem anteversion, manual methods may produce considerable error.