SU-FF-T-22: A Method to Reduce the Dose Uncertainty Caused by High Energy Cutoffs for Monte Carlo Treatment Planning

Abstract

Purpose: A method to reduce the statistical uncertainty of dose caused by high energy cutoffs for electron transport was implemented in our homegrown Monte Carlotreatment planning system. Method and Materials: In Monte Carlo radiation transport, an electron is discarded and its energy is deposited locally when its total energy is below a cutoff energy. The depositedenergy is significantly higher than that calculated using the CSDA model with the corresponding restricted stopping powers. This will create a higher statistical uncertainty on dose and generate a confusing dose distribution, especially when low‐density voxel exists. In this work, a new technique was developed by continuously transporting a discarded electron without considering electron multiple scattering or secondary particle generation. It has a continuous energy loss based on its mass collision stopping power in the local medium with an additional energy loss (about 70%) to account for the effect of approximations made in transporting the electron in a straight line rather than a curved path. Results: After the new method was applied, the statistical uncertainties of the doses in air cavities of a head‐and‐neck patient was reduced from up to 39% to the same level of that in the surrounding tissue which is only about 2%. The dose statistical uncertainties of the tissue voxels were also reduced by 9% of their initial values. The simulation time with the new method was increased by 9%. And thus, the simulation efficiency was increased by 9% when the energy cutoff is 0.7MeV. When a cutoff of 1.5MeV was used, the new method increased the simulation efficiency by a factor of 3. Conclusion: A new technique was developed to reduce the statistical uncertainty of doses in low‐density voxels caused by high energy cutoffs for electron transport. The calculation efficiency and the dose distributions were improved significantly.