DVH-Based Inverse Planning Using Monte Carlo Dosimetry for LDR Prostate Brachytherapy

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Imaging

Background:

• Modern low-dose-rate brachytherapy planning relies on the TG-43 formalism, which can lead to suboptimal plans due to dose overestimation.
• Existing clinical systems often do not fully utilize total dose information, potentially compromising treatment efficacy.

Purpose of the Study:

• To develop an advanced inverse planning approach for low-dose-rate brachytherapy.
• To enhance treatment plan quality and precision by integrating dose-volume histogram (DVH) data.
• To maintain clinically acceptable planning times without sacrificing accuracy.

Main Methods:

• Utilized precomputed Monte Carlo (MC) dose kernels for accurate dosimetry calculations.
• Implemented graphics processing unit (GPU) acceleration to reduce MC computational burden for intraoperative use.
• Developed a DVH-based fast simulated annealing optimization algorithm.

Main Results:

• Generated treatment plans in an average of 37.5 ± 3.2 seconds.
• Achieved comparable prostate dose coverage to existing methods.
• Demonstrated improved prostate dose homogeneity (up to 6.1%) and reduced urethral dose (up to 4.0%).

Conclusions:

• A novel DVH-based optimization algorithm incorporating MC dosimetry was successfully developed.
• The integration of DVH requirements provided enhanced control over treatment plan optimization.
• Considering tissue heterogeneity further improved the quality of the optimized treatment plans.