Knowledge-based inverse treatment planning for low-dose-rate prostate brachytherapy.

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Biology

Background:

• Permanent low-dose-rate brachytherapy is a standard prostate cancer treatment.
• Treatment planning for brachytherapy is complex and requires significant expertise.
• Current inverse planning methods can yield suboptimal plans, necessitating time-consuming manual adjustments.

Purpose of the Study:

• To develop and evaluate a novel knowledge-based (KB) optimization method for prostate cancer brachytherapy planning.
• To generate clinically acceptable treatment plans efficiently, minimizing manual intervention.
• To reduce overall treatment planning time and potentially operating room duration.

Main Methods:

• A KB model was created using dose distributions from previous clinical brachytherapy plans.
• Plans were optimized using a constraint inverse optimization problem mimicking template dose plans.
• The method was benchmarked against current clinical practice in a retrospective study, evaluating optimization time, dosimetric performance, and clinical acceptability.
• A Turing test assessed the quality of the KB model's generated plans.

Main Results:

• The KB model incorporated five expert-selected, high-quality treatment plans.
• Treatment plans for 20 patients were generated rapidly, with optimization times ranging from 6 to 29 seconds.
• Dosimetric performance was equivalent to current clinical practice, as confirmed by statistical analysis (Wilcoxon signed rank test).
• The Turing test indicated that KB-generated plans were equivalent to those from current practice.

Conclusions:

• The proposed KB model effectively captures user-specific dose features for generating acceptable brachytherapy plans.
• Acceptable treatment plans can be created with a single optimization run in under a minute.
• This approach has the potential to significantly reduce operating room time and patient anesthesia duration.