通过能量矩阵正则化进行质子点扫描弧形治疗的能量层优化。
Energy layer optimization via energy matrix regularization for proton spot-scanning arc therapy.
机构信息
University of Michigan-Shanghai Jiao Tong University Joint Institute, Shanghai Jiao Tong University, Shanghai, China.
Department of Radiation Oncology, University of Kansas Medical Center, Kansas City, Missouri, USA.
出版信息
Med Phys. 2022 Sep;49(9):5752-5762. doi: 10.1002/mp.15836. Epub 2022 Jul 25.
PURPOSE
Spot-scanning arc therapy (SPArc) is an emerging proton modality that can potentially offer a combination of advantages in plan quality and delivery efficiency, compared with traditional IMPT of a few beam angles. Unlike IMPT, frequent low-to-high energy layer switching (so called switch-up (SU)) can degrade delivery efficiency for SPArc. However, it is a tradeoff between the minimization of SU times and the optimization of plan quality. This work will consider the energy layer optimization (ELO) problem for SPArc and develop a new ELO method via energy matrix (EM) regularization to improve plan quality and delivery efficiency.
METHODS
The major innovation of EM method for ELO is to design an EM that encourages desirable energy-layer map with minimal SU during SPArc, and then incorporate this EM into the SPArc treatment planning to simultaneously minimize the number of SU and optimize plan quality. The EM method is solved by the fast iterative shrinkage-thresholding algorithm and validated in comparison with a state-of-the-art method, so-called energy sequencing (ES).
RESULTS
EM is validated and compared with ES using representative clinical cases. In terms of delivery efficiency, EM had fewer SU than ES with an average of 35% reduction of SU. In terms of plan quality, compared with ES, EM had smaller optimization objective values and better target dose conformality, and generally lower dose to organs-at-risk and lower integral dose to body. In terms of computational efficiency, EM was substantially more efficient than ES by at least 10-fold.
CONCLUSION
We have developed a new ELO method for SPArc using EM regularization and shown that this new method EM can improve both delivery efficiency and plan quality, with substantially reduced computational time, compared with ES.
目的
点扫描弧形治疗（SPArc）是一种新兴的质子治疗模式，与少数射束角度的传统调强质子治疗（IMPT）相比，它具有潜在的优势，能够在计划质量和治疗效率方面实现优势互补。与 IMPT 不同，频繁的低至高能量层切换（所谓的“上切”（SU））会降低 SPArc 的治疗效率。然而，这是在最小化 SU 次数和优化计划质量之间的权衡。本研究将考虑 SPArc 的能量层优化（ELO）问题，并通过能量矩阵（EM）正则化开发一种新的 ELO 方法，以提高计划质量和治疗效率。
方法
EM 方法用于 ELO 的主要创新之处在于设计一个 EM，鼓励在 SPArc 期间以最小的 SU 获得理想的能量层图，然后将其纳入 SPArc 治疗计划中，以同时最小化 SU 的次数并优化计划质量。通过快速迭代收缩阈值算法解决 EM 方法，并与一种称为能量排序（ES）的最先进方法进行验证。
结果
通过使用代表性的临床病例验证并比较 EM 和 ES。在治疗效率方面，EM 比 ES 的 SU 少，平均减少了 35%的 SU。在计划质量方面，与 ES 相比，EM 的优化目标值更小，靶区剂量适形性更好，并且通常危及器官的剂量更低，全身积分剂量更低。在计算效率方面，EM 比 ES 至少快 10 倍。
结论
我们已经开发了一种使用 EM 正则化的 SPArc 的新 ELO 方法，并表明与 ES 相比，这种新的 EM 方法可以提高治疗效率和计划质量，同时大大减少计算时间。
Zotero 插件