TLD measurements of 6MV x-ray dose in a 2-D phantom for breast treatment

Full text: Planning Computers and their Algorithms have been used in Radiotherapy Departments in some form since the late sixties, (Milan and Bentley, BJR ,1973). The intention of this project is to determine if the algorithms, used in the Planning Computer to obtain dose distribution data, succeed in satisfying the limits set by ICRU report 24, i.e: ± 5%, when applied to a 2-D Phantom in the shape of a simple breast. External Contour Correction, and an Inhomogeneity Correction, Equivalent Tissue Air Ratio (ETAR) were tested. Dose plans were generated on a G.E. Target Series 2 Planning Computer. Dose levels were compared in a cylindrical solid water phantom which contained a lung insert for the purpose of testing inhomogeneity corrections. To measure the absorbed dose at the specific points TLD chips of LiF:Mg,Ti (3.1x3.1x0.9mm3) were used. The doses were delivered by a Varian Clinac 2100C linac. Four distinct treatment plans, relating to typical treatment regimes for breast therapy, were generated using a 6 MV photon beam and a 10x10 cm2 field size. The first three dose plans, mimicked a 2-D solid water phantom being dosed with no, 2cm and 4cm overshoots in order to analyse curvature corrections undertaken by the algorithms. The fourth plan allowed the tissue inhomogeneity correction to be tested by allowing a lung insert to be placed in the phantom with TLD chips placed at strategic points to record dose delivered at these points. Results indicate that treatment plans undertaken in the homogeneous breast phantom yield doses with differences greater than 5%, results involving the lung insert versus ETAR and Monte Carlo are also reported. The data gathered and the analyses undertaken led to the conclusions that the correction algorithms and dose planning systems are not as accurate as desired. Given the need in radiotherapy planning for dose distributions to be developed as a compromise between speed and accuracy the external contour correction is done to adequate levels, however care should be exercised in the interpretation of inhomogeneity corrected results