Abstract：
Objective.Dynamic trajectory radiotherapy (DTRT) and dynamic mixed-beam arc therapy (DYMBARC) exploit non-coplanarity and, for DYMBARC, simultaneously optimized photon and electron beams. Margin concepts to account for set-up uncertainties during delivery are ill-defined for electron fields. We develop robust optimization for DTRT&DYMBARC and compare dosimetric plan quality and robustness for both techniques and both optimization strategies for four cases.Approach.Cases for different treatment sites and clinical target volume (CTV) to planning target volume (PTV) margins,m, were investigated. Dynamic gantry-table-collimator photon paths were optimized to minimize PTV/organ-at-risk (OAR) overlap in beam\'s-eye-view and minimize potential photon multileaf collimator (MLC) travel. For DYMBARC plans, non-isocentric partial electron arcs or static fields with shortened source-to-surface distance (80 cm) were added. Direct aperture optimization (DAO) was used to simultaneously optimize MLC-based intensity modulation for both photon and electron beams yielding deliverable PTV-based DTRT&DYMBARC plans. Robust-optimized plans used the same paths/arcs/fields. DAO with stochastic programming was used for set-up uncertainties with equal weights in all translational directions and magnitudeδsuch thatm= 0.7δ. Robust analysis considered random errors in all directions with or without an additional systematic error in the worst 3D direction for the adjacent OARs.Main results.Electron contribution was 7%-41% of target dose depending on the case and optimization strategy for DYMBARC. All techniques achieved similar CTV coverage in the nominal (no error) scenario. OAR sparing was overall better in the DYMBARC plans than in the DTRT plans and DYMBARC plans were generally more robust to the considered uncertainties. OAR sparing was better in the PTV-based than in robust-optimized plans for OARs abutting or overlapping with the target volume, but more affected by uncertainties.Significance.Better plan robustness can be achieved with robust optimization than with margins. Combining electron arcs/fields with non-coplanar photon trajectories further improves robustness and OAR sparing.
摘要：
目的：动态轨迹放射治疗（DTRT）和动态混合束电弧治疗（DYMBARC）利用非共面性，对于DYMBARC，同时优化光子和电子束。对于电子场，考虑交付过程中设置不确定性的余量概念定义不明确。我们为DTRT和DYMBARC开发了鲁棒优化，并比较了两种技术和四种情况下优化策略的剂量测定计划质量和鲁棒性。
方法。病例针对不同的治疗部位和临床目标体积（CTV）到计划目标体积（PTV）的边缘，m,被调查了。动态机架-台-准直器光子路径进行了优化，以最大程度地减少PTV/危险器官（OAR）在光束眼视图中的重叠，并最大程度地减少潜在的光子多叶准直器（MLC）传播。对于DYMBARC计划，添加了具有缩短的源至表面距离（80厘米）的非等中心部分电子弧或静态场。直接孔径优化（DAO）用于同时优化光子和电子束的基于MLC的强度调制，从而产生可交付的基于PTV的DTRT和DYMBARC计划。稳健优化计划使用相同的路径/弧/场。具有随机规划的DAO用于在所有平移方向和幅度δ上具有相等权重的设置不确定性，使得m=0.7δ。稳健分析考虑了所有方向上的随机误差，在相邻OAR的最差3D方向上有或没有额外的系统误差。
主要结果。取决于情况和DYMBARC的优化策略,电子贡献为目标剂量的7%-41%。所有技术在标称(无错误)情况下实现了类似的CTV覆盖。与DTRT计划相比，DYMBARC计划中的OAR节省总体上更好，而DYMBARC计划通常对所考虑的不确定性更为稳健。基于PTV的OAR节约比与目标体积相邻或重叠的OAR的稳健优化计划更好，但更受不确定性影响。
意义。与裕度相比，通过鲁棒优化可以实现更好的计划鲁棒性。将电子弧/场与非共面光子轨迹组合进一步提高了鲁棒性和OAR节省。
