Evaluation of the ray-casting analytical algorithm for pencil beam scanning proton therapy

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Dosimetry

Background:

• Analytical dose calculation engines are standard in pencil beam scanned (PBS) proton therapy planning.
• Recent studies question the suitability of these analytical methods for PBS treatments.
• Key analytical approaches include ray-casting (RC) and pencil-beam (PB) methods.

Purpose of the Study:

• To compare dose distributions and dosimetric indices from RC and PB methods against Monte Carlo (MC) calculations.
• To evaluate the impact of dose grid resolution on calculation accuracy.
• To assess the clinical relevance of discrepancies between analytical and MC methods.

Main Methods:

• Comparison of dose distributions and dosimetric indices using clinical PBS treatment plans.
• Analysis performed on clinical dose calculation grids and varying dose grid resolutions.
• Validation against comprehensive Monte Carlo (MC) simulations.

Main Results:

• Pencil-beam (PB) calculations with double grid resolution showed the highest voxel agreement (98.6%/94.0%) within ±5%/±3% compared to MC.
• Despite point-to-point dose differences, clinically relevant dosimetric indices for PTV and organs at risk agreed within ±4% and ±5%, respectively.
• Ray-casting (RC) offered a lower computational cost than PB.

Conclusions:

• Analytical methods show discrepancies with MC, but clinically relevant dosimetric indices remain in good agreement.
• Ray-casting (RC) is suitable for rapid dose optimization in PBS proton therapy.
• Combining RC optimization with MC for absolute dose scaling and plan quality assurance is recommended.