C Fellner1, B Sommer, N Siedhoff, R Pötter. 1. Universitätsklinik für Strahlentherapie und -biologie, AKH Wien. claudia.fellner@akh-wien.ac.at
Abstract
BACKGROUND: Modern 3-dimensional treatment planning systems on the basis of sectional imaging allow the calculation of volumes for organs of interest. The aim of this study is to investigate systematically the accuracy of calculations of volumes by the use of a phantom for 4 different treatment planning systems. MATERIAL AND METHOD: The tests were done with a phantom with 5 cylindrical structures and 6 spherical shaped structures. After performing a CT-scan and reading the data into the planning systems the structures were contoured and the volumes were calculated in order to compare these values with the values calculated by mathematical equations. This was systematically done as a function of different parameters. RESULTS: Comparing different methods of contouring showed notable influence on the result. Parameters as number of calculation points or length of cylinders showed no significant differences. In summary, the mean deviations for cylinders were +7% for system A, -2% for B, -17% for C, and 0% for D. For larger spheres (radii between 5 and 2.5 cm) the mean deviations were -5% for A, +3% for B, +1% for C, and +5% for D. For smaller spheres (radii between 1.75 and 1.25 cm) the mean deviations were -14% for A, -2% for B, -10% for C, and -4% for D. CONCLUSION: Verifying results of planning systems is important for the daily routine, but it has to be taken into account, that small changes of the radius of a cylinder or sphere cause substantial volume changes. The differences are also caused by inaccuracies of the whole procedure, e.g., the CT study, the shape and dimensions of the cylinders and the spheres and the CT information and the delineariation of the structures.
BACKGROUND: Modern 3-dimensional treatment planning systems on the basis of sectional imaging allow the calculation of volumes for organs of interest. The aim of this study is to investigate systematically the accuracy of calculations of volumes by the use of a phantom for 4 different treatment planning systems. MATERIAL AND METHOD: The tests were done with a phantom with 5 cylindrical structures and 6 spherical shaped structures. After performing a CT-scan and reading the data into the planning systems the structures were contoured and the volumes were calculated in order to compare these values with the values calculated by mathematical equations. This was systematically done as a function of different parameters. RESULTS: Comparing different methods of contouring showed notable influence on the result. Parameters as number of calculation points or length of cylinders showed no significant differences. In summary, the mean deviations for cylinders were +7% for system A, -2% for B, -17% for C, and 0% for D. For larger spheres (radii between 5 and 2.5 cm) the mean deviations were -5% for A, +3% for B, +1% for C, and +5% for D. For smaller spheres (radii between 1.75 and 1.25 cm) the mean deviations were -14% for A, -2% for B, -10% for C, and -4% for D. CONCLUSION: Verifying results of planning systems is important for the daily routine, but it has to be taken into account, that small changes of the radius of a cylinder or sphere cause substantial volume changes. The differences are also caused by inaccuracies of the whole procedure, e.g., the CT study, the shape and dimensions of the cylinders and the spheres and the CT information and the delineariation of the structures.