TU-G-BRB-03: Iterative Optimization of Normalized Transmission Maps for IMRT Using Arbitrary Beam Profiles

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Biology

Background:

• Flattening filter free (FFF) beams offer higher dose rates but possess non-uniform profiles.
• Existing inverse planning methods struggle with arbitrary beam profiles, limiting treatment optimization.
• Accurate modeling of beam profiles is crucial for maximizing therapeutic efficacy.

Purpose of the Study:

• To develop a novel dose optimization scheme incorporating inherent beam profiles for arbitrary beam shapes.
• To enhance the utilization of flattening filter free (FFF) beam properties in radiation therapy planning.
• To preserve the convexity of the optimization problem while maximizing treatment efficacy.

Main Methods:

• Decomposed fluence maps into inherent beam profiles and normalized transmission maps (NTMs).
• Developed a least-squares problem constrained by NTM total-variation for optimal fluence maps.
• Iteratively optimized and renormalized NTMs and beam profiles in a closed loop.

Main Results:

• Demonstrated improved dose conformity in head-neck and prostate cancer cases.
• The novel method achieved more conformal dose distributions compared to existing solutions.
• Preserved piecewise constancy in dose distributions while enhancing delivery efficiency.

Conclusions:

• The proposed formalism offers a unified framework for inverse planning with arbitrary and mixed beam profiles.
• Total-variation constraints on NTMs enable optimal balancing of dose conformity and deliverability.
• This approach maximizes the benefits of FFF beams and improves treatment planning accuracy.