X-ray beam modelling in radiotherapy: the effect of lung inhomogeneities

Abstract

A lung phantom consisting of epoxy resin based analogs is developed in-house and a quality assurance method, which compares experimental CT numbers with theoretical CT numbers calculated from electronic cross sections, is described. Currently used photon beam inhomogeneity correction models and their inadequacies are discussed. The methods studied include Batho and Equivalent Tissue-Air-Ratio corrections. For a 10 MV high energy photon beam the mean difference in central axis depth dose in the lung phantom for these methods is 8.8% and 3.5%, respectively (for a 5 x 5 cm field). The differences in the beam profiles at an off-axis distance of 5 cm is 12.6% for both methods (for a 10 x 10 cm field). A unified three-dimensional superposition approach to dose calculations used in treatment planning of a polyenergetic 10 MV photon beam in radiotherapy is developed. This radiotherapy X-ray beam computation method involves an electron gamma shower (EGS) Monte Carlo generated surface polyenergetic dose spread array (PDSA), which describes the energy spread from a point interaction source. This is superposed with the relative reduction in polyenergetic total energy released per unit mass (TERMA) as the beam traverses tissue. By comparing primary PDSAs produced at different radiological depths, the effect of beam hardening on the PDSA has been quantified. Calculations show the mean electron range due to the surface 10 MV primary PDSA is 6.67 mm and the mean electron range of the beam hardened primary PDSA is 8.24 mm. In comparison a 3 MV primary monoenergetic dose spread array (MDSA) has a much smaller mean electron range of 4.81 mm. The effect of using a beam hardened PDSA for superposition is also studied. The mean percentage difference between depth dose curves obtained using superposition of a surface and a beam hardened PDSA is only 0.1 %. The mean percentage difference from experimental data for these superposition curves is 1.2% down to 20 cm in a homogeneous phantom. The superposition process is shown to be forgiving to spectral differences in the PDSA but sensitive to spectral changes in the TERMA. A method of scaling PDSAs to account for lung inhomogeneities is introduced which shows good agreement with experimental and Monte Carlo results in the lung phantom. The mean difference in central axis depth dose in the lung phantom for this method is 2.0% for a 5 x 5 cm field. The differences in lung for the beam profile at an off-axis distance of 5 cm is 4.3% for a 10 x 10 cm field.