Therapeutic treatment plan optimization with probability density-based dose prescription.

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Biology

Background:

• Conventional inverse planning uses rigid dose prescriptions, often leading to ill-conditioned optimization problems.
• Existing methods lack flexibility in handling variations in desired dose delivery.
• Clinical experience and planner priorities are difficult to incorporate directly into optimization.

Purpose of the Study:

• To propose a generalized dose optimization scheme using statistical formalism and preference functions.
• To enable more flexible dose prescription and improve the conditioning of the optimization problem.
• To provide a framework for incorporating planner's priorities and clinical experience into inverse planning.

Main Methods:

• Developed a dose optimization scheme based on statistical formalism, replacing rigid doses with preference functions.
• Utilized an iterative dose optimization algorithm for system optimization.
• Studied performance using hypothetical C-shaped tumor and prostate cases.

Main Results:

• The proposed preference function approach allows flexible manipulation of dose distributions.
• Optimization problems become less ill-defined by allowing deviation from the most desired dose.
• The method successfully generated optimized dose distributions aligned with planner specifications.

Conclusions:

• The new framework effectively formalizes planner priorities and incorporates them into dose optimization.
• Preference functions offer enhanced control over treatment plans, facilitating Intensity-Modulated Radiation Therapy (IMRT) planning.
• This approach represents a more general and adaptable method for radiation therapy dose optimization.