Minimum MU optimization (MMO): an inverse optimization approach for the PBS minimum MU constraint.

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Biology

Background:

• Proton pencil beam scanning (PBS) treatment planning faces challenges with minimum monitor unit (MU) constraints, impacting dose delivery accuracy.
• Ensuring the deliverability of planned doses is crucial for effective radiation therapy outcomes.

Purpose of the Study:

• To introduce and validate a novel inverse optimization approach, minimum MU optimization (MMO), to address the minimum MU constraint in PBS planning.
• To enhance the accuracy of deliverable proton therapy dose distributions.

Main Methods:

• Formulation of the minimum MU problem as an inverse optimization problem (MMO).
• Application of iterative convex relaxations to handle the non-convex nature of MMO.
• Development of an alternating direction method of multipliers (ADMM) algorithm to solve convex subproblems.

Main Results:

• The proposed MMO approach successfully enforces minimum MU constraints.
• Comparison with the greedy reassignment (GR) algorithm showed MMO yields more accurate deliverable plans, as indicated by improved γ-index results.
• A practical ADMM-based MMO was developed for high-quality PBS treatment planning.

Conclusions:

• The ADMM-based MMO provides an effective solution for the minimum MU constraint in PBS.
• This method improves the accuracy and deliverability of proton therapy treatment plans.
• MMO represents a significant advancement for precise radiation oncology delivery.