Automated rapidplan model validation using Eclipse scripting API

使用 Eclipse 脚本 API 实现 RapidPlan 模型的自动化验证

阅读:1

Abstract

RapidPlan offers efficiency gains and quality improvements in treatment planning. Prior to its use in the clinic, it requires an extensive validation procedure in which established clinical plans and those generated by the model are compared. The manual iterative nature of this process is resource intensive, as numerous iterations are required to fine-tune and optimize a RapidPlan model. To streamline the efficiency and reduce the resource burden of RapidPlan model validation, a standalone executable auto planning script was written in C# leveraging the Eclipse scripting application programming interface (ESAPI). The script automatically batch generates treatment plans, as well as exports and plots the population-based DVH metrics of these plans, without any user input, reducing the time and effort required to explore and refine model objectives. Configured with adjustable parameters via Excel and .txt files, this approach allows end users to change input variables quickly and easily without needing to re-approve the script. The script has been implemented for a variety of treatment sites, including intact prostate, prostate & nodes, lung, rectum, unilateral head & neck, bilateral head & neck, and liver stereotactic body radiation therapy (SBRT). The process for liver SBRT has been used here as an example to illustrate the use and power of the script. Over numerous iterations, 76 patients in the model set and 17 patients in a validation set were replanned using the script, creating a total of 405 automatic plans with an overall active planning time of 118.7 h. This study demonstrates the effectiveness of automating the RapidPlan model validation process, significantly reducing the time and resource burden associated with traditional manual methods.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。