| 曹瑞芬,李国丽,宋钢,等.用于逆向放疗计划多目标优化的改进快速非支配排序遗传算法ANSGA-Ⅱ[J].中华放射医学与防护杂志,2007,27(5):467-470.CAO Rui-fen,LI Quo-li,SONG Gang,et al.An improved fast and elitist multi-objective genetic algorithm-ANSGA-Ⅱ for multi-objective optimization of inverse radiotherapy treatment planning[J].Chin J Radiol Med Prot,2007,27(5):467-470 |
| 用于逆向放疗计划多目标优化的改进快速非支配排序遗传算法ANSGA-Ⅱ |
| An improved fast and elitist multi-objective genetic algorithm-ANSGA-Ⅱ for multi-objective optimization of inverse radiotherapy treatment planning |
| 投稿时间：2006-12-05 |
| DOI： |
| 中文关键词: 逆向计划 多目标优化 多目标进化优化算法 NSGA-Ⅱ ANSGA-Ⅱ |
| 英文关键词:Inverse planning Multi-objective optimization Multi-objective evolutionary optimization algorithm NSGA-Ⅱ |
| 基金项目:国家“973”计划项目(2006CB708307),安徽省自然科学基金项目(070413081) |
|
| 摘要点击次数: 3679 |
| 全文下载次数: 2465 |
| 中文摘要: |
| 目的 对逆向放射治疗计划进行多目标优化算法研究,以期为逆向放射治疗计划系统提供高效的优化算法选择。方法 快速非支配排序遗传算法-NSGA-Ⅱ作为多目标进化优化算法的代表,优于其他算法。利用NSGA-Ⅱ在多目标优化中的优势,并针对其交叉变异操作不够灵活,进行改进形成自适应交叉变异快速非支配排序遗传算法(简称ANSGA-Ⅱ);寻优过程中,根据逆向放射治疗计划优化的特性,在算法每一代充分利用决策变量的先验知识进行种群生成,来提高算法的全局寻优能力。结果 用优化一张人体头部CT片上靶区、危及器官、其他正常组织的平均剂量实例进行测试,本文算法可以在几分钟内找到满意解。结论 本文算法可以为实际的逆向放射治疗计划优化提供优化方法选择。 |
| 英文摘要: |
| Objective To provide a fast and effective multi-objective optimization algorithm for inverse radiotherapy treatment planning system. Methods Non-dominated Sorting Genetic Algorithm-NSGA-Ⅱ is a representative of multi-objective evolutionary optimization algorithms and excels the others. The paper produces ANSGA-Ⅱ that makes use of advantage of NSGA-Ⅱ, and uses adaptive crossover and mutation to improve its flexibility; according the character of inverse radiotherapy treatment planning, the paper uses the pre-known knowledge to generate individuals of every generation in the course of optimization, which enhances the convergent speed and improves efficiency. Results The example of optimizing average dose of a sheet of CT,including PTV、OAR、NT, proves the algorithm could find satisfied solutions in several minutes. Conclusions The algorithm could provide clinic inverse radiotherapy treatment planning system with selection of optimization algorithms. |
| HTML 查看全文 查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|