End-to-end testing of automatic plan optimization using RayStation scripting for hypofractionated multimetastatic...

Area of Science:

• Radiation Oncology
• Medical Physics
• Radiosurgery

Background:

• External beam stereotactic radiosurgery for multiple brain metastases presents challenges in selecting optimal treatment isocenters due to unique patient-specific target volumes and critical structure proximity.
• Current methods can be time-consuming and may not achieve optimal dose distribution.

Purpose of the Study:

• To compare isocenter-optimized plans generated by RayStation's Python scripting with clinically used Pinnacle plans.
• To validate the accuracy and efficiency of RayStation for treatment planning and delivery in stereotactic radiosurgery for multiple brain metastases.

Main Methods:

• Ten patient plans were automatically regenerated in RayStation using Python scripts to generate four scenario-based isocenter placements.
• Plans were compared to Pinnacle plans using dose-volume metrics (V4Gy, V12Gy), dose to the brainstem, homogeneity and conformity indices, gradient index, and monitor units.
• End-to-end testing with phantoms was performed for quality assurance.

Main Results:

• RayStation plans demonstrated statistically significant improvements in homogeneity and conformity indices, and a more favorable gradient index.
• Normal brain dose was reduced, with a 13% decrease in V12Gy and a 5% decrease in V4Gy.
• Plan optimization time was reduced by 64%, and total monitor units per fraction decreased by 20%.

Conclusions:

• Scenario-based automatic plan generation using RayStation's Python scripting effectively identifies optimal treatment isocenters.
• RayStation plans offer superior quality, particularly in reducing normal brain dose, compared to Pinnacle plans.
• RayStation is a suitable modality for hypofractionated stereotactic radiosurgery of multiple brain metastases.