A multi-objective optimization method for seed distribution in prostate cancer low-dose-rate brachytherapy.

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Biology

Background:

• Brachytherapy planning aims to cover tumors with prescribed radiation dose while sparing normal organs.
• Current inverse planning methods often use subjective single objective functions, leading to suboptimal treatment plans.
• Optimizing seed implantation is crucial for effective prostate low-dose-rate brachytherapy (LDRBT).

Purpose of the Study:

• To develop and evaluate a novel seed implantation position optimization algorithm for prostate LDRBT.
• The goal is to enhance dose distribution and meet stringent clinical dosimetric criteria.
• To improve upon existing inverse planning methods by addressing limitations in handling multiple objectives.

Main Methods:

• A multi-objective non-dominated genetic algorithm was employed to optimize seed implantation positions.
• The algorithm directly optimized dose-volume indices and integrated inverse planning parameters.
• The proposed method was tested on a prostate case and compared against single-objective optimization.

Main Results:

• Single-objective optimization failed to satisfy all four clinical dose constraints in tested cases.
• The multi-objective optimization approach yielded a Pareto front with superior solutions.
• 24.8% of multi-objective plans met some conditions, and six plans fully satisfied all constraints, balancing urethral dose.

Conclusions:

• Multi-objective optimization provides a more comprehensive set of solutions for prostate LDRBT.
• This approach effectively balances conflicting clinical goals, outperforming single-objective methods.
• The proposed algorithm demonstrates superior performance in achieving clinically acceptable treatment plans.