Methods:

Three spherical tantalum markers of 1.0 mm in diameter were inserted into three peripheral pegs of an all polyethylene glenoid component. The glenoid component was implanted in a sawbone scapula. To determine the effect of metal artifact on quantification of glenoid implant motion, two sawbone humerii were used: one without the prosthetic humeral components and the other with the prosthetic humeral head and stem. Three custom-made translucent spacers with the uniform thickness were placed between the glenoid component and the scapula to produce a gradual translation of the glenoid component from 1 mm to 3 mm. Before and after inserting each spacer, the surface of the glenoid component was digitized by a MicroScribe. The surface points were used to fit a sphere and the corresponding center of the sphere was calculated. The actual translation of the glenoid component was measured as the three-dimensional (3D) distance between the center of the sphere before and after insertion of each spacer. Then, the shoulder model was scanned by a clinical CT with and without the spacers for both humerii conditions. Velcro straps were used to secure the humerus to the glenoid component between the trials. All CT scans were reconstructed in VolNinja software to superimpose the scapula positions (Figure 1). The three tantalum markers were visualized and the center coordinates of the markers were used to measure the 3D distance before and after insertion of each spacer. The accuracy was defined by the difference between the averaged 3D distance measured by CT reconstruction and that measured by the MicroScribe. The standard deviation of the 3D distance measured by each tantalum marker was calculated to evaluate the reliability of the tantalum marker visualization.