Background
In this study the accuracy of a pencil beam based treatment planning system (TPS) was evaluated for lung dose calculations by comparison with measurement and the Monte Carlo (MC) method.
Aim
In the current study we assessed the performance of the Eclipse treatment planning system in the thorax region by ionization chamber measurements and Monte Carlo calculations. We examined two analytic methods: modifi ed Batho (MB) and equivalent tissue-air ratio "ETAR" methods for thorax region irradiations. For Monte Carlo calculations in the thorax phantom, we modelled a Varian Clinac 2100EX linac. After benchmarking our model with water phantom measurements we used this model for thorax phantom calculations.
Materials/Methods 8 and 15MV photon beams of Varian 21EX linac were used for irradiations. Using MANP4C Monte Carlo code, the geometry of the linac head was simulated. After commissioning "MC" beam models, lung doses were calculated by the Monte Carlo (MC) method. Irradiation cases were: (1) posterior fi elds of single lung with fi eld sizes of 4×4 and 10×10cm 2 (2) lateral fi elds of thorax with 4×4 and 10×10cm 2 fi eld sizes.
Results
TPS calculations involving ETAR and MB methods were in close agreement with
Monte Carlo results and measurements for a 10×10 cm 2 fi eld size at both energies. For a fi eld size of 4×4cm 2 the maximum differences in local dose between TPS calculations and measurement were +33% (MB) and +28% (ETAR). Also, they ignored lung dose reduction due to lateral electronic equilibrium for small fi eld size. Similar results would be expected for other TPSs implementing these algorithms. MC calculations were in excellent agreement with measurement, showing local differences of no more than 2% for all measured points.
Conclusions
Our study fi ndings showed great differences between both analytical methods and measurements for 4×4cm 2 fi eld sizes for points in the lung. Our study recommends using the MC method for small-fi eld lung dose calculations.