Clinical application of locally trained and re-optimized lung SBRT RapidPlan model
10.13491/j.issn.1004-714X.2025.05.016
- VernacularTitle:肺癌立体定向放射治疗自动计划模型本地训练及再优化后的临床应用
- Author:
Libo ZHANG
1
;
Weiqiang GE
1
;
Ye YAO
1
Author Information
1. Department of Radiotherapy, Huadong Hospital, Fudan University, Shanghai 200040, China.
- Publication Type:OriginalArticles
- Keywords:
Lung cancer;
Stereotactic body radiation therapy;
RapidPlan
- From:
Chinese Journal of Radiological Health
2025;34(5):726-732
- CountryChina
- Language:Chinese
-
Abstract:
Objective To explore the clinical effectiveness of a locally trained and re-optimized RapidPlan model in lung stereotactic body radiation therapy (SBRT). Methods A total of 132 lung SBRT cases were retrospectively collected as a training group for the establishment of an initial RapidPlan model (RP-1) in the Eclipse 15.5 treatment planning system. RP-1 was then used to optimize the training group plans to generate the first-generation RapidPlans, which were used as a new training group to generate the upgraded RapidPlan model (RP-Plus). Additional 40 lung SBRT cases were selected to verify the quality and efficiency of RP-Plus plans, which were further assessed through clinical quality assurance (QA). Results The RP-Plus reduced the average number of optimization cycles by 2.1 times and 0.3 times compared to manual and RP-1 methods, respectively, and reduced the optimization time by 57.6% and 18.8%. The conformity index met clinical requirements, and the gradient index and Max2cm(%) were improved compared to RP-1 (both P < 0.05). The RP-Plus plans showed lower dosimetric parameters of organs at risk, including lung V20Gy, spinal cord Dmax, and heart \begin{document}$D_{0.1{\mathrm{cm}}^3} $\end{document} The clinical QA pass rate of the RP-Plus plans was 100%. Conclusion The locally trained and re-optimized lung SBRT RapidPlan model can further improve plan quality and efficiency, and meet clinical QA requirements, showing feasibility for clinical application.
