摘要
目的:利用Perl语言编写Pinnacle计划系统脚本的嵌入式自动化程序,进行多分次鼻咽癌调强治疗计划快速自动设计,研究多分次自动计划设计的可行性及设计效率。方法:选取14例鼻咽癌患者,每例患者均在整个治疗过程中接受4次CT模拟图像(分别为CT1、CT2、CT3、CT4)。在Pinnacle计划系统(9.2m版本)上基于首次CT模拟图像(CT1)手工设计出满足处方剂量和危及器官剂量限制要求的调强放疗计划,并将CT1上手工计划的目标函数设置参数用Perl语言写入Pinnacle计划系统脚本,应用于同一患者的后续3个分次间CT图像(CT2、CT3、CT4)的自动计划设计。同时,由工作经验分别为不足1年和大于5年的两位物理师基于CT2独立设计手工计划,标记为Ma和Mb组。自动计划与手工计划的设计时间进行配对t检验。另外,由一名工作经验为10年以上的临床放疗医师评估手工计划和自动计划的临床可行性。结果:在CT2、CT3和CT4上设计的自动计划时间与Ma、Mb组的手工计划时间分别为(3.358±0.490)、(2.837±0.340)、(2.754±0.270)、(43.169±8.980)、(39.088±13.210)min,可应用于临床的计划所占比例分别为78.6%、78.6%、71.4%、78.6%、78.6%,且自动计划比手动计划在时间上有明显减少(P<0.05)。结论:利用Perl语言基于第一分次的目标函数编写的自动化程序,可以完成后续多分次鼻咽癌自动计划设计。运用自动计划能够节省计划人员大量的计划设计时间,保证计划的设计质量,并为鼻咽癌多分次快速放疗计划设计提供一种便捷方案。
Objective To evaluate the feasibility and efficiency of multi-fractionated autoplanning on nasopharyngeal carcinoma(NPC)with Perl-based scripts in Pinnacle treatment planning system.Methods A total of 14 NPC patients were selected,and 4 sets of simulated CT images(CT1,CT2,CT3 and CT4)for each patient were derived during the treatment.Each radiotherapy plan on its first set of simulated CT images(CT1)was manually created according to the prescribed dose to target areas and dose constraints to organs-at-risk on Pinnacle treatment planning system(version 9.2m),while the setting of objective functions on the manual plan were read with Perl into scripts of Pinnacle planning system and applied for autoplanning on the 3 successive sets of CT images(CT2,CT3 and CT4)of the same patient.Manual plans based on CT2 were executed by 2 radiation physicists with experiences less than 1 year and more than 5 years,and the plans were grouped into Ma and Mb.The paired t test was used to evaluate the time for manual planning and autoplanning.In addition,the feasibility of manual planning and autoplanning were evaluated by an experienced radiation oncologist with experience over 10 years.Results For the automated plans based on CT2,CT3,CT4 and manual plans from Ma and Mb,the planning times were(3.358±0.490),(2.837±0.340),(2.754±0.270),(43.169±8.980),(39.088±13.210)min,respectively,and the percentage of the number of clinically available plans were 78.6%,78.6%,71.4%,78.6%,78.6%,respectively.The autoplanning significantly reduced planning time compared to manual planning(P<0.05).Conclusions The automated plans based on successive sets of CT images for NPC can be achieved using the autoprogramming based on the first set of CT images with Perl-based scripts.The application of autoplanning not only saves planners plenty of time but also guarantees the quality of planning,providing an easier and quicker approach for multi-fractionated radiotherapy planning for NPC.
作者
程品晶
翁星
戴艺涛
柏朋刚
李奇欣
陈济鸿
曹叶
CHENG Pinjing;WENG Xing;DAI Yitao;BAI Penggang;LI Qixin;CHEN Jihong;CAO Ye
出处
《中国医学物理学杂志》
CSCD
2018年第12期1386-1391,共6页
Chinese Journal of Medical Physics
基金
福建省自然科学基金(2015Y0010)
福建省卫生计生中青年骨干人才培养项目(2017-ZQN-15)
关键词
鼻咽癌
调强放疗
自动计划
目标函数
nasopharyngeal carcinoma
intensity-modulated radiotherapy
autoplanning
objective function