Multi-objective optimization in radiotherapy: applications to stereotactic radiosurgery and prostate brachytherapy

Area of Science:

• Medical Physics
• Computational Biology
• Radiotherapy

Background:

• Radiation therapy planning is a complex multi-objective optimization challenge.
• Existing methods often require significant expert time and may not achieve optimal outcomes.

Purpose of the Study:

• To develop and evaluate a machine intelligent scheme for radiation therapy treatment planning.
• To integrate multi-objective decision analysis (MODA) with genetic algorithm (GA) optimization for enhanced planning.

Main Methods:

• A novel MODA scheme utilizing L(p) metric and a dynamic gauge function was developed.
• The MODA scheme was coupled with a genetic algorithm (GA) for adaptive optimization.
• The approach was tested on stereotactic radiosurgery for brain lesions and prostate brachytherapy under uncertainty.

Main Results:

• GA-optimized plans for brain lesions demonstrated superior sparing of critical normal tissues compared to human-developed plans.
• The machine optimization produced novel strategies and significantly reduced planning time.
• The MODA-GA approach effectively managed noisy objectives in prostate brachytherapy, simulating surgical uncertainties.

Conclusions:

• The combination of MODA and GA offers a powerful solution for practical radiation therapy planning.
• This intelligent scheme has the potential for real-time applications and can complement expert knowledge.
• The method shows promise in handling uncertainties and improving treatment plan quality and efficiency.