Can the Student Outperform the Master? A Plan Comparison Between Pinnacle Auto-Planning and Eclipse knowledge-Based...

Area of Science:

• Radiation Oncology
• Medical Physics
• Computational Biology

Background:

• Automated planning engines are crucial for optimizing intensity-modulated radiotherapy (IMRT).
• Pinnacle Auto-Planning uses iterative optimization, while Eclipse RapidPlan employs knowledge-based prediction.
• Understanding their comparative performance is essential for clinical adoption.

Purpose of the Study:

• To objectively compare the performance of Pinnacle Auto-Planning and Eclipse RapidPlan.
• To evaluate their efficacy in IMRT planning for prostate fossa and lymphatics.
• To assess plan quality using the 2011 American Association of Medical Dosimetrists Plan Challenge metric.

Main Methods:

• An identical IMRT beam setup and simultaneous integrated boost prescription were used for all plans.
• Auto-Planning created initial plans, which were used to build a RapidPlan model.
• Both engines planned a test set of 10 patients and the Plan Challenge case without manual intervention.

Main Results:

• No significant difference in overall performance was observed between Auto-Planning and RapidPlan (P = .509).
• Most of the 14 quantitative submetrics showed no significant difference between the two engines.
• On the Plan Challenge case, both engines scored highly, comparable to average human planner performance.

Conclusions:

• Pinnacle Auto-Planning and Eclipse RapidPlan demonstrate comparable performance in automated IMRT planning.
• Both systems generate treatment plans equivalent to those produced by experienced human planners.
• RapidPlan's performance was found to be as good as, but not superior to, its library plans.