Technical Report: Development and Implementation of an Open Source Template Interpretation Class Library for...

Area of Science:

• Radiation oncology
• Medical physics
• Software development

Background:

• Automated treatment planning in radiation therapy is challenging due to clinic and physician variability.
• Consistent plan parameters are difficult to achieve with manual planning.

Purpose of the Study:

• To develop an open-source class library to enhance efficiency and consistency in automated radiation therapy planning.
• To overcome variability in clinical preferences for automated treatment planning.

Main Methods:

• Developed an open-source class library to interpret clinical templates for automated planning.
• Leveraged the library for automated planning of 39 patients.
• Retrospectively compared automated plans with 78 clinically approved manual plans.

Main Results:

• Successfully generated 74 of 78 plans without manual intervention.
• Automated plans showed improved target dose homogeneity (HI: 3.11 vs 3.30).
• Significantly reduced generalized equivalent uniform dose (gEUD) in femurs and rectum (P ≤ 0.001).
• Automated plans passed quality assurance gamma analysis (>95%) and demonstrated deliverability.

Conclusions:

• Automated planning for prostate cancer and pelvic node radiation therapy is feasible using volumetric modulated arc therapy (VMAT) and clinical templates.
• The developed class library effectively interfaces between templates and planning systems for automated, high-quality plan generation.
• Standardized clinical workflows enable the automation of high-quality radiation therapy plans.