Date of Award
Master of Engineering Science
Thomas R. Jenkyn
A markerless radiostereometric analysis (RSA) system was developed to measure three- dimensional (3D) skeletal kinematics using biplanar fluoroscopy. A virtual set-up was created, in which the fluoroscope foci and image planes were positioned. Computed tomography (CT) was used to create 3D bone models that were imported into the virtual set-up and manually moved until their projections, as viewed from the two foci, matched the two images. The accuracy of the markerless RSA system in determining relative shoulder kinematic translations and orientations was evaluated against the “gold standards” of a precisions cross-slide table and a standard RSA system, respectively. Average root mean squared errors (RMSEs) of 0.082 mm and 1.18° were found. In an effort to decrease subject’s radiation exposure, the effect of lowering CT dosage on markerless RSA accuracy was evaluated. Acceptable accuracies were obtained using bone models derived from one-ninth of the normal radiation dose.
Allen, Anne-Marie V., "Development and Validation of a Markerless Radiostereometric Analysis (RSA)System" (2009). Digitized Theses. 3934.