Model-based Roentgen stereophotogrammetric analysis (RSA) using a Novel Radiopaque UHMWPE

Roentgen stereophotogrammetric analysis (RSA) is a technique to assess the position of implants with respect to their surrounding environment with a sub-millimetre accuracy for metallic implants. However, due to the low visibility of polyethylene implants, RSA cannot currently be used for polymeric implants. This study aim was to investigate the use of a novel radiopaque UHMWPE for RSA and its precision level. To create radiopaque UHMWPE Oxford Unicompartmental knee arthroplasty (OUKA) bearings (n=4, Zimmer-Biomet, UK) were treated with a contrast agent (Lipiodol, Guerbert, France, T= 105°C for 0,12,18, and 24h). A standard RSA set-up (two synchronized X-ray tubes positioned at 1.5 m above the film cassette with an angle of 20°) was used to acquire stereo-radiographs of 10 successive poses. A carbon uniplanar calibration box (Leiden, Netherlands) was positioned underneath the phantom. Translational and rotational movements in X-, Y- a Zdirection were applied to mimic clinical positioning variation. Zero motion between the markers and the implants was assumed, hence the measured migrations by the RSA software included the experimental errors. Secondly, a known amount of separation was applied (0.02 to 2 mm, n=5) using a micrometer stage to the phantom by lifting-up the femoral component. The separation distance was calculated using the RSA software and compared with the micrometer measurements. The results showed radiopaque UHMWPE can be used for RSA analysis with a precision level comparable with the metallic component for any translational movements (mean: Tx= 0.06 mm Ty=0.03 mm, Tz= 0.06mm). However, this level is marginally higher for any rotational movement (mean: Rx= 0.069°Ry=0.224°, Rz= 0.089°). KruskalWallis tests showed no difference between the different level of radiopacity or the direction of the measurement (p=0.3 and 0.05 respectively). Furthermore, measured lift-up can be calculated with sub-millimetre accuracy (absolute average mean error = 0.01 mm). In conclusion, this study used a novel type of polyethylene and confirmed the possibility of using this polymer for RSA studies.