Acuros External Beam (Acuros XB) is a new radiation dose calculation algorithm available as part of Varian Medical Systems' radiotherapy treatment planning system, ECLIPSE. Acuros XB calculates dose distributions by finding the deterministic solution to the linear Boltzmann transport equation which governs the transport of particles or radiation through matter. Among other things, Acuros XB claims an ability to accurately model dose perturbations due to increased photon and electron scatter within a high-density volume, such as a hip prosthesis. Until now, the only way to accurately model high-density scatter was with a Monte Carlo simulation which gives the stochastic solution to the same transport equation, but is time and computationally expensive. In contrast, Acuros XB solves the transport equation at time scales appropriate for clinical use. An evaluation of Acuros XB for radiation dose calculations involving high-density objects was undertaken using EGSnrc based Monte Carlo as the benchmark. Calculations were performed for geometrically ideal virtual phantoms, water tank phantoms containing cylindrical steel rods and hip prostheses, and for a clinical prostate treatment plan involving a unilateral prosthetic hip. The anisotropic analytical algorithm (AAA), a convolution-superposition algorithm used for treatment planning at the British Columbia Cancer Agency's Vancouver Island Center, was also used to illustrate the limitations of current radiotherapy planning tools. In addition, to verify the qualitative properties of dose perturbations due to high-density volumes, film measurements were taken and compared to Monte Carlo, Acuros XB and AAA data. Dose distributions calculated with Acuros XB agree very well with distributions calculated with Monte Carlo. Gamma-analyses performed at 2% and 2 mm using Monte Carlo as the reference dose were within tolerance for 92-99% of voxels considered. AAA, on the other hand, was within tolerance for 61-97% of voxels considered under the same gamma-constraints. For the clinical prostate plan, AAA produced localized dose underestimates that were absent when calculated by Acuros XB. As well, both Monte Carlo and Acuros XB showed very good agreement with the film measurements, while AAA showed large discrepancies at and beyond the location of measured dose perturbations. Acuros XB has been shown to handle does perturbations due to high-density volumes as well as Monte Carlo, at clinically appropriate time scales, and better than the current algorithm used for treatment planning at the Vancouver Island Center. / Graduate

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第20248号 / 医博第4207号 / 新制||医||1020(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 鈴木 実, 教授 別所 和久, 教授 大森 孝一 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM