Comparing 2 Monte Carlo Systems in Use for Proton Therapy Research

The effect of voxelization in Monte Carlo simulation to validate Bragg peak characteristics for a pencil proton beam

Background

The purpose of this research was to show how the Bragg peak (BP) characteristics were affected by changing the voxel size in longitudinal and transverse directions in Monte Carlo (MC) simulations by using Geant4 and to calculate BP characteristics accurately by considering the voxel size effect for 68 MeV and 235.81 MeV.

Materials and methods

Different interpolation techniques were applied to simulation data to find the closest results to the experimental data.

Results

When the x-size of the voxel was increased 2 times at low energy, the maximum dose increase in the entrance and plateau regions were 17.8% and 17%, respectively, while BP curve shifted to the shallower region, resulting in a 0.5 mm reduction in the curable tumor width (W_80pd). At high energy, the maximum dose increase at the entrance and plateau regions were 0.4% and 0.6%, respectively, while it was observed that W_80pd did not change. When the y-z sizes of the voxel were increased 2 times at low energy, the maximum dose reduction at the entrance and plateau regions was 3.4%, but no change was observed in W_80pd. At high energy, when the y-z sizes of the voxel were increased 2.2 times, the maximum dose reduction at the entrance and plateau regions were 8.9% and 9.1%, respectively, while W_80pd increased by 0.5 mm. When linear, cubic spline, and Akima interpolations were applied to the simulation data, it was found that the results closest to the experimental data were obtained for Akima interpolations for both energies.

Conclusion

it has been shown that the voxel size effect for the longitudinal direction was more effective at low energy than at high energy. However, the voxel size effect for the transverse direction was more effective for high energy.

Simulation of spread-out bragg peaks in proton beams using Geant4/TOPAS

The simulation of proton Spread-Out Bragg Peaks (SOBPs) was implemented using the Geant4-based TOPAS Monte Carlo software. Dynamic proton energy switching was implemented using TOPAS time features, while beam weights were calculated using an empirical power law formalism with Bragg peaks spaced by 0.5 mm. To find power parameters yielding flat SOBPs we sampled power parameters for maximum kinetic energies of 50 MeV to 250 MeV and SOBP widths of 15% to 40% of the depth of the distal SOBP end. Simulations were run in a 50 cm cubic water phantom using a uniform squared proton beam. Depth dose was scored along the central axis in a binned cylinder with 1 cm diameter in 2.5 mm increments. Power parameters yielding a flat SOBPs were found to vary with, both energy and SOBP width and differed significantly from previously reported values based on simulations with MCNPX.