O3271 RADIOSTEREOMETRY OF MARKER-LESS HIP PROSTHESES, A COMPARISON OF TWO MODEL BASED METHODS

Abstract

Aims Intro: Radiostereometry (RSA) is a powerful tool for assessment of prosthesis migration in total hip replacement. Typically marker beads are inserted into the bone and into turrets inserted into the prosthesis. Where the prosthesis can not be beaded valuable information may be lost. We present two methods that determine the three-dimensional prosthesis position using geometric models. Method: The þrst method approximates the stem of the prosthesis to an axially symmetric cylinder and spherical head. The method is fast but may lead to errors due to the cylinder approximation. The second uses accurate three-dimensional surface models of each implant model. Clinical precision was determined from repeat same day RSA examinations. Results: Given as 99% precision limit rotation ranged from 0.4 degrees (sagittal) to 2.5 degrees (axial) for the cylinder method (n = 38), compared to 0.2 degrees (sagittal) to 0.6 degrees (axial) for the 3d model (n = 6). Translations were similar ranging from 0.2mm axial to 0.7mm sagittal. Conclusion: Model based RSA can give useful results where beads can not be implanted into the prosthesis. Although the primary method gives good results these are less precise than that for beaded RSA. Simulations infer a potential accuracy for the method of a few microns, however this is currently limited by the closeness of correspondence of an individual stem to its model. The method here is applicable to any implant type. Accurate 3D models are required, however for new designs these should be available.