基于启发式优化的切线式乳腺癌调强放疗自动计划

Automated planning of tangential breast intensity-modulated radiotherapy using heuristic optimization.

机构信息

Radiation Medicine Program, Princess Margaret Hospital, University Health Network, Toronto, ON, Canada.

出版信息

Int J Radiat Oncol Biol Phys. 2011 Oct 1;81(2):575-83. doi: 10.1016/j.ijrobp.2010.11.016. Epub 2011 Jan 13.

Abstract

PURPOSE

To present an automated technique for two-field tangential breast intensity-modulated radiotherapy (IMRT) treatment planning.

METHOD AND MATERIALS

A total of 158 planned patients with Stage 0, I, and II breast cancer treated using whole-breast IMRT were retrospectively replanned using automated treatment planning tools. The tools developed are integrated into the existing clinical treatment planning system (Pinnacle(3)) and are designed to perform the manual volume delineation, beam placement, and IMRT treatment planning steps carried out by the treatment planning radiation therapist. The automated algorithm, using only the radio-opaque markers placed at CT simulation as inputs, optimizes the tangential beam parameters to geometrically minimize the amount of lung and heart treated while covering the whole-breast volume. The IMRT parameters are optimized according to the automatically delineated whole-breast volume.

RESULTS

The mean time to generate a complete treatment plan was 6 min, 50 s ± 1 min 12 s. For the automated plans, 157 of 158 plans (99%) were deemed clinically acceptable, and 138 of 158 plans (87%) were deemed clinically improved or equal to the corresponding clinical plan when reviewed in a randomized, double-blinded study by one experienced breast radiation oncologist. In addition, overall the automated plans were dosimetrically equivalent to the clinical plans when scored for target coverage and lung and heart doses.

CONCLUSION

We have developed robust and efficient automated tools for fully inversed planned tangential breast IMRT planning that can be readily integrated into clinical practice. The tools produce clinically acceptable plans using only the common anatomic landmarks from the CT simulation process as an input. We anticipate the tools will improve patient access to high-quality IMRT treatment by simplifying the planning process and will reduce the effort and cost of incorporating more advanced planning into clinical practice.

摘要

目的

介绍一种用于二维切线野乳腺癌调强放疗(IMRT)治疗计划的自动化技术。

方法和材料

回顾性地对 158 例接受全乳 IMRT 治疗的 0 期、I 期和 II 期乳腺癌患者进行了计划重新制定,使用的是自动化治疗计划工具。所开发的工具集成到现有的临床治疗计划系统(Pinnacle(3))中,旨在执行由治疗计划放射治疗师进行的手动容积勾画、射束放置和 IMRT 治疗计划步骤。自动化算法仅使用 CT 模拟时放置的不透射线标记作为输入,优化切线射束参数,使治疗的肺和心脏组织最小化,同时覆盖整个乳房体积。IMRT 参数根据自动勾画的全乳体积进行优化。

结果

生成完整治疗计划的平均时间为 6 分 50 秒±1 分 12 秒。对于自动化计划,在随机、双盲研究中,由一位经验丰富的乳腺癌放射肿瘤学家进行评估,158 例计划中有 157 例(99%)被认为是临床可接受的,其中 138 例(87%)被认为是临床改善或与相应的临床计划相当。此外,在目标覆盖、肺和心脏剂量评分方面,总体而言,自动化计划与临床计划在剂量学上是等效的。

结论

我们已经开发出了用于全反式计划切线野乳腺癌 IMRT 计划的强大而高效的自动化工具,可以很容易地集成到临床实践中。该工具仅使用 CT 模拟过程中的常见解剖学标志作为输入,即可生成临床可接受的计划。我们预计这些工具将通过简化计划过程,为更多患者提供高质量的 IMRT 治疗,并减少将更先进的计划纳入临床实践的努力和成本。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索