Validating a double Gaussian source model for small proton fields in a commercial Monte-Carlo dose calculation...

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Dosimetry

Background:

• Proton pencil beams produce secondary particles (spray) that can impact dose calculations in small fields.
• Inaccurate modeling of this spray in treatment planning systems (TPS) can lead to dose distribution errors.
• The Monte Carlo (MC) engine in RayStation TPS requires accurate beam modeling for precise dosimetry.

Purpose of the Study:

• To benchmark the RayStation TPS Monte Carlo dose engine in small proton fields.
• To compare single Gaussian (SG) and double Gaussian (DG) models for representing initial proton fluence and nozzle spray.
• To improve the accuracy of dose calculations in small proton fields.

Main Methods:

• Proton fluence distributions were measured using a scintillation screen.
• Single Gaussian (SG) and double Gaussian (DG) models were fitted to measured profiles.
• Scan-field factors (SFs) and point doses in spherical targets were measured in water and compared to TPS predictions.

Main Results:

• Single Gaussian (SG) modeling showed deviations >2% in fields <4x4cm², up to 5.8%.
• Directly fitted double Gaussian (DG) modeling reduced deviations to <2% for all field sizes and energies.
• Point dose deviations were reduced from 3.3% (SG) to 2.0% (DG).

Conclusions:

• A double Gaussian (DG) beam model accurately represents nozzle spray in RayStation's MC engine.
• This DG model reduces deviations in small spherical targets to below 2%.
• Uncertainty analysis indicates the combined standard uncertainty of measurements is comparable to the achieved accuracy.