3D coordinate computation by RSA under non-standard conditions

The operative requirements of Roentgen Stereophotogrammetric Analysis (RSA) change according to the type of investigation being performed, and sometimes practical conditions do not permit to respect standard specifications. The aim of this work is to verify the reliability of the system under nonstandard conditions for both the hip and knee calibration cages. Different focus-to-film distances, angles of convergence between the two X-ray tubes, and slopes of the radiographic cassettes were investigated. A test object containing five markers, was X-rayed in the cage reference system under the different experimental conditions, and the images obtained were measured following the RSA criteria. Two way analysis of variance was applied to each marker coordinate to evaluate the fluctuation caused by repeated measurements of the same couple of images, and different experimental conditions. The statistical analysis showed that no parameter represented a significant source of variation. In particular, the negligible variation of the 3D coordinates due to focus-to-film distance change. demonstrated that the system can correctly compute the marker position inside the calibrated space under all conditions of magnifications. The fluctuation of marker coordinates due to the different parameters, was equal or less than 10/sup -4/ mm, strictly within the limits required by clinical investigations. These results permit a reliable use of the RSA technique In new applications, in which particular tools could be required to simulate mechanical stress occurring during daily life. Thus, this study contributes to improving the flexibility of RSA in clinical studies, making the technical requirements of the radiographic set-up less constraining.