Adaptive step size algorithm to increase efficiency of proton macro Monte Carlo dose calculation.

Area of Science:

• Medical Physics
• Computational Dosimetry
• Radiation Oncology

Background:

• Proton radiation therapy requires accurate and efficient dose calculations in complex voxelized geometries.
• Current Monte Carlo (MC) methods can be computationally intensive, limiting their clinical applicability.
• The proton macro Monte Carlo (pMMC) method offers a balance between accuracy and speed but can be further optimized.

Purpose of the Study:

• To enhance the speed and accuracy of proton dose calculations in voxelized environments for proton therapy.
• To implement and evaluate an adaptive step size algorithm within the proton macro Monte Carlo (pMMC) framework.

Main Methods:

• Extended the local-to-global MMC method with an adaptive step size algorithm for efficient proton transport.
• Utilized a pre-simulated database for sampling transport parameters and adaptively determined slab sizes based on material interfaces.
• Validated the algorithm against non-adaptive pMMC and full MC simulations using various beam types, energies, and phantoms, including a head and neck patient CT.

Main Results:

• Adaptive step size selection, particularly nearest neighbor slab thickness, optimized the trade-off between accuracy and efficiency.
• Lateral adaptive reduction of slab size was necessary for material interfaces closer than 0.5 mm.
• Achieved dose differences within 1% or 1 mm compared to full Geant4 MC simulations, with efficiency gains up to 5.6x (academic phantoms) and 3.4x (patient CT) over non-adaptive pMMC, and significantly higher gains over Geant4.

Conclusions:

• The implemented adaptive step size algorithm for pMMC delivers dose calculation accuracy comparable to full MC simulations.
• Demonstrated substantial efficiency gains, achieving up to a factor of three improvement over the non-adaptive pMMC and two orders of magnitude over full MC for complex patient data.
• This adaptive algorithm represents a significant advancement for fast and accurate proton therapy dose calculations in clinical settings.