Geometric Asymmetry and Rotation In Proton Pbs Spots: Dosimetric Sensitivity to Beam Model Parameterization
Abstract
Purpose
Proton pencil beam scanning (PBS) spot profiles are traditionally modeled using orthogonal X and Y profiles. In-air measurements may exhibit asymmetry and rotation. This study evaluates whether rotation-aware elliptical spot characterization produces meaningful dosimetric differences compared to conventional symmetric beam model parameterization.
Methods
In-air proton spot profiles were measured using a high-resolution 2D scintillator detector (0.2 mm) for 227.1 MeV at seven longitudinal positions relative to isocenter (±30, ±20, ±10 cm, and isocenter). Spot boundaries were fitted to an ellipse using direct least-squares conic method via singular value decomposition, yielding major axis, minor axis, and rotation angle. Lateral profiles along major and minor axes were compared with conventional orthogonal X–Y profiles in the clinical beam model. The beam model used TPS automodeling workflow, enforcing symmetric Gaussian spot representation (σx = σy). Dosimetric impact was evaluated in a water phantom using point dose, lateral, and depth dose profiles. Square-field sensitivity was assessed using a maximum-energy layer and three spread-out Bragg peak (SOBP) fields. Four anthropomorphic phantom test plans (Brain, Esophagus, Prostate, and Spine) were evaluated.
Results
Spots exhibited rotation angles from 22.3° to 29.3°. Elliptical fitting yielded major and minor spot widths differing from the symmetric σ enforced by automodeling, reflecting features not explicitly modeled. Dosimetric impact was minor. For square SOBP fields, point dose differences were approximately 0.3%, with close agreement in lateral and depth profiles. Across composite plans, target DVH and point dose differences were mostly within 1%. Localized differences up to ~3% occurred for some organ-at-risk D1 values in high-gradient regions, without systematic shifts in target coverage.
Conclusion
Rotation-aware spot characterization reveals substantial asymmetry suppressed by current symmetric beam models. Spot overlap, multiple Coulomb scattering, and energy-layer averaging mitigate spot-level differences, contribute to explaining PBS dose calculation robustness while highlighting commercial beam model parameterization limitations.