调强放射治疗（Intensity Modulated Radiation Therapy IMRT)是一种先进的三维适形放 射治疗技术，它的出现给放射肿瘤学带来了一次重大变革。 者设强定疗的向目划标统，以利据用划优定化放算计法可自计动调系根制逆寻找最优的入射角度、布野方式和强度分布，使计划的制定过程自动化。 在项目组前期在调强放疗的剂量计算、多目标优化方法等关键问题研究的基础上，本 报告针对 IMRT 多目标逆向计划系统进行设计、开发，为项目组后期的研究工作提供平台 支持。 报告的主要工作有， （1）对系统涉及的多种优化算法、剂量计算方法进行了研究，对不同算法间的接口 数据、文件格式进行了定义； （2）对系统进行了坐标定义，使不同剂量计算程序集成到系统时有章可循； （3）对系统的逆向计划过程进行了流程定义，该流程根据剂量计算的特点，避免了 优化过程中重复调用剂量计算程序，大大节省了优化时间； （4）对设计的多目标逆向计划系统的各功能模块进行了阐述； （5）通过具体算例验证了系统的可行性。 报告的研究工作得到国家自然基金（60872112）、浙江省自然基金项目（Z207588）的 共同资助。 关键词，IMRT，逆向计划系统，多目标优化浙江工业大学硕士学位报告 ii THE DESIGN AND FEASIBILITY ANALYSIS ON THE MULTI-OBJECTIVE INVERSE PLANNING SYSTEM OF IMRT ABSTRACT IMRT (Intensity Modulated Radiation Therapy IMRT) is an advanced three-dimensional conformal radiation technology. It brings a revolution for radiation oncology. Using optimization algorithms, IMRT inverse planning system finds the optimal angle of incidence, fields and intensity distribution according to prearrange target, which makes the planning process automatic. Based on the previously work of our group about multi-objective optimizations, dose calculation algorithms and some other key issues, this paper focus on the design, development and feasibility analysis of the multi-objective inverse planning system, provides the support platform for further research of our group. The primary work of the thesis is as fellows: （1）Avariety of optimization and dose calculation algorithms integrated into the system are studied. The interface data between different algorithms, file formats are defined; （2）The coordinate system is defined, so that the different dose calculation algorithm integrated into the system is not disorderly; （3）The flow of inverse planning is defined, which according to the characteristics of dose calculation algorithm, avoiding re-call the dose calculation procedure for each optimization generation, significant saving inverse planning time; （4）The functional modules of multi-objective inverse planning system are described. （5）The feasibility of the system is verified by the test examples.浙江工业大学硕士学位报告 iii The National Natural Science Foundation Project of China (60872112) & Zhejiang Province Natural Science Foundation Project (Z207588) support the research. Key Words: IMRT, Inverse treatment planning system, Multi-objective optimization浙江工业大学硕士学位报告 目 录 摘要................................................................... I 第 1 章 绪 论........................................................... 1 1.1 概述............................................................................ 1 1.2 调强放射治疗.................................................................... 1 1.2.1 调强放射治疗概述 .......................................................................... 1 1.2.2 调强放射治疗的发展历史 .................................................................... 2 1.2.3 调强放射治疗的逆向计划 .................................................................... 2 1.2.4 调强放射治疗的实现方式 .................................................................... 3 1.2.5 调强放射治疗计划系统的国内外研究现状及关键技术 ............................................ 4 1.3 报告的主要内容 ................................................................. 7 第 2 章 逆向计划系统中的多目标优化算法 .................................... 8 2.1 算法基础........................................................................ 8 2.1.1 遗传算法 .................................................................................. 8 2.1.2 模拟退火算法 ............................................................................. 11 2.2 多目标优化的基本概念........................................................... 13 2.3 多目标优化算法................................................................. 14 2.3.1 多目标遗传算法 ........................................................................... 15 2.3.2 基于交叉变异扰动的MOSA算法 ............................................................... 16 2.3.3 改进的快速非支配分类遗传算法 ............................................................. 19 2.3.5 多目标梯度算法 ........................................................................... 23 2.5 小结........................................................................... 24 第 3 章 IMRT多目标逆向计划系统的设计 ..................................... 26 3.1 概述........................................................................... 26 3.2 系统分析....................................................................... 26 3.2.1 系统的设计目标 ........................................................................... 26 3.2.2 系统设计的关键技术 ....................................................................... 26 3.3 IMRT多目标逆向计划系统设计 .................................................... 33 3.3.1 系统模块组成 ............................................................................. 33 3.3.2 逆向计划的过程 ........................................................................... 33 3.4 系统各主要模块的功能及用户界面................................................. 36 3.5 小结........................................................................... 39 第 4 章 IMRT多目标逆向计划系统可行性研究 ................................. 40 4.1 算例一......................................................................... 40浙江工业大学硕士学位报告 4.2 算例二......................................................................... 43 4.3 算例三......................................................................... 45 4.4 小结 .......................................................................... 48 第 5 章 结论与展望 ....................................................... 49 5.1 结论........................................................................... 49 5.2 展望........................................................................... 49