SU‐E‐T‐676: Dosimetric Dependence on Variations of the Lung Density and Geometry: A Monte Carlo Evaluation Using Virtual Dynamic Heterogeneous Phantom

Abstract

Purpose: This Monte Carlo study investigates the dosimetry of the lung—soft tissue interface under different lung geometries using virtual dynamic heterogeneous phantoms. Methods: 6 and 18 MV photon beams (field sizes = 4×4 and 10×10 cm2), produced by a Varian 21EX linac were used. Three lung phantoms namely slab, column and cube representing different lung geometries were irradiated by the photon beams. The lung volume and density were varied in five phases mimicking the lung motion in breathing. The lung density was set to 0.15, 0.2, 0.25, 0.3 and 0.35 g/cc corresponding to phases 1 to 5 with the lung volume changed while keeping the lung mass constant. Relative depth doses and beam profiles for each phantom in the five phases were calculated using the DOSXYZnrc. Results: Our results show that there is obvious dosimetric variation around the lung—soft tissue interface among the five phases in the three lung phantoms. This is due to the positional and density change of the interface. Interpolation was done to register all dose distributions in the five phases to one in order to compare with the dosimetry of a static lung phantom (e.g. only phase 3 with lung density = 0.25 g/cc). Dose deviation was found to be about ±5%, when a breathing lung with changes of volume and density was considered. Moreover, the deviation is more significant for photon beams of relatively higher energy (18 MV) and smaller field size (4×4 cm2). Conclusions: Through Monte Carlo studies of lung heterogeneous correction, we concluded that more realistic dosimetry may be acquired by considering a virtual dynamic phantom to mimic the breathing motion. The accuracy of the lung motion can further be improved by increasing the number of phase and applying weight factor to each phase in the breathing cycle.