A novel fluence map optimization model incorporating leaf sequencing constraints.

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Biology

Background:

• Current radiation therapy planning models struggle to balance fluence map optimization with leaf sequencing constraints.
• Existing smoothing models often compromise treatment efficiency for plan quality.

Purpose of the Study:

• To introduce a novel fluence map optimization model that incorporates leaf sequencing constraints directly.
• To balance fluence map and leaf sequencing optimization by adding total monitor unit (TNMU) requirements to model constraints.

Main Methods:

• Formulated new models for synchronized and non-synchronized leaf sweeping schemes (left-to-right).
• Transformed models into linear constrained quadratic programming problems solvable by the interior point method.
• Evaluated models using head-neck and prostate cancer clinical treatment datasets.

Main Results:

• The proposed models significantly outperformed recent smoothing models (total variance and quadratic smoothing).
• Objective dose deviation increased slower with decreasing TNMU in the new models.
• Achieved better TNMU reduction while maintaining similar plan conformity.

Conclusions:

• The novel model effectively integrates leaf sequencing constraints into fluence map optimization.
• This approach offers improved efficiency and plan quality in radiation therapy.
• The method provides a superior balance between optimization objectives and treatment delivery constraints.