Keller Harry, Jaffray David A, Rosewall Tara, White Elizabeth
Department of Radiation Physics, Princess Margaret Hospital, Toronto, Ontario M5G 2M9, Canada.
Med Phys. 2006 May;33(5):1388-97. doi: 10.1118/1.2192580.
With the introduction of image-guided radiation therapy (IGRT) delivery systems on-line set-up correction strategies have gained in popularity. Usually, the correction workload of these strategies is high compared to off-line strategies as daily setup corrections have to be performed based on a predefined action level. In this work, it is proposed that on-line strategies must not only be judged in terms of workload but also in terms of efficacy. While workload can be easily predicted for such strategies, the efficacy must ultimately reflect the efficiency with which the original treatment plan intent is met. The purpose of this work is to investigate the tradeoff between workload and efficacy of three different on-line set-up correction strategies: The common fixed action level strategy and two novel on-line setup correction strategies, i.e., a dose-volume histogram (DVH) constraint and an equivalent uniform dose (EUD) score strategy that aim directly for better compliance with original treatment plan intent. All strategies were reformulated in terms of a score function that reflected treatment plan intent. A retrospective study was conducted on 5 prostate patients (7-field conformal, 79.8 Gy, 42 fractions). PTV margins were 10 mm except in the posterior direction (7 mm). The original treatment plan intent for these patients was defined using a set of DVH constraints. The results show that the on-line setup correction strategy based on a fixed action level of 3 mm resulted in a considerable correction workload. For larger action levels, a dose benefit (in terms of EUD) in the rectum and bladder was observed for all patients which is clinically "fortuitous" but difficult to take advantage of. In contrast, the application of the two novel strategies generally resulted in a controlled decrease of the dose to the rectum and the bladder with a smaller workload. It is concluded that using information about target anatomy and the planned dose distribution allows the design of specific correction strategies that are better tailored to the individual patient and that comply effectively with initial treatment plan intent.
随着图像引导放射治疗（IGRT）输送系统的引入，在线设置校正策略越来越受欢迎。通常，与离线策略相比，这些策略的校正工作量较高，因为必须根据预定义的行动水平进行每日设置校正。在这项工作中，有人提出，在线策略不仅要根据工作量来判断，还要根据疗效来判断。虽然此类策略的工作量很容易预测，但疗效最终必须反映满足原始治疗计划意图的效率。这项工作的目的是研究三种不同的在线设置校正策略在工作量和疗效之间的权衡：常见的固定行动水平策略以及两种新颖的在线设置校正策略，即剂量体积直方图（DVH）约束和等效均匀剂量（EUD）评分策略，它们直接旨在更好地符合原始治疗计划意图。所有策略都根据反映治疗计划意图的评分函数进行了重新制定。对5名前列腺癌患者进行了回顾性研究（7野适形，79.8 Gy，42次分割）。除后方方向（7 mm）外，计划靶体积（PTV）边缘为10 mm。这些患者的原始治疗计划意图是使用一组DVH约束来定义的。结果表明，基于3 mm固定行动水平的在线设置校正策略导致了相当大的校正工作量。对于更大的行动水平，所有患者的直肠和膀胱均观察到剂量获益（以EUD计），这在临床上是“偶然的”，但难以利用。相比之下，两种新颖策略的应用通常导致直肠和膀胱剂量的可控降低，且工作量较小。得出的结论是，利用有关靶区解剖结构和计划剂量分布的信息，可以设计出更适合个体患者且能有效符合初始治疗计划意图的特定校正策略。
