Optimization of target grouping in distributive stereotactic radiosurgery using the excel evolutionary solver.

Area of Science:

• Radiation Oncology
• Medical Physics
• Computational Biology

Background:

• Distributive stereotactic radiosurgery (dSRS) involves fractionating treatment for multiple metastases across different days.
• Optimizing target groupings in dSRS is crucial for maximizing spatial separation and treatment safety.
• Current methods for target grouping may not be systematically optimized.

Purpose of the Study:

• To develop and validate an accessible optimization technique for distributing brain metastases into optimal treatment fractions using a genetic algorithm.
• To enhance spatial separation between targets treated in the same fraction during dSRS.

Main Methods:

• Utilized the Evolutionary Solver in Excel to optimize target volume groupings for dSRS.
• Tested the algorithm with geometric test cases, random simulations, and clinical GammaKnife patient data.
• Employed an objective function maximizing average inter-target distances with constraints on targets per fraction and minimum separation.

Main Results:

• The Evolutionary Solver successfully identified optimal target groupings in all test cases.
• Optimized groupings increased mean inter-target distance by 9% and minimum distance by 57% compared to random groupings.
• Clinical cases showed improved mean distances (81.6 to 85.6 mm) and minimum separation (35.2 to 51.6 mm) with the optimized method.

Conclusions:

• The Evolutionary Solver in Excel offers a systematic and reproducible approach for optimizing dSRS target groupings.
• This method enhances spatial separation, potentially improving treatment outcomes and safety in dSRS.
• The developed technique is accessible and applicable to clinical dSRS planning.