Comparison between Monte Carlo simulation and measurement with a 3D polymer gel dosimeter for dose distributions in biological samples

In this research, we used a 135 MeV/nucleon carbon-ion beam to irradiate a biological sample composed of fresh chicken meat and bones, which was placed in front of a PAGAT gel dosimeter, and compared the measured and simulated transverse-relaxation-rate (R-2) distributions in the gel dosimeter. We experimentally measured the three-dimensional R-2 distribution, which records the dose induced by particles penetrating the sample, by using magnetic resonance imaging. The obtained R-2 distribution reflected the heterogeneity of the biological sample. We also conducted Monte Carlo simulations using the PHITS code by reconstructing the elemental composition of the biological sample from its computed tomography images while taking into account the dependence of the gel response on the linear energy transfer. The simulation reproduced the experimental distal edge structure of the R-2 distribution with an accuracy under about 2 mm, which is approximately the same as the voxel size currently used in treatment planning.