Comparison of AAA and Acuros Xb for Small‑Field Lung SBRT Dosimetry
Abstract
Purpose
To evaluate the accuracy of the Analytical Anisotropic Algorithm (AAA) and Acuros XB (AXB) implemented in the Eclipse treatment planning system for small-field dosimetry in low-density lung, with emphasis on their performance in lung stereotactic body radiation therapy (SBRT).
Methods
Field output factors (FOFs) from 0.5×0.5 cm² to 10×10 cm² were computed in a lung phantom at 10 cm depth (2.5 cm solid water and 7.5 cm cork) using a 6 MV TrueBeam beam commissioned with a minimum field size of 2×2 cm², compared against EGSnrc Monte Carlo simulations. Six SBRT plans were created for spherical targets of 1.5 cm and 2.0 cm in diameter with overridden lung densities of 0.2, 0.3, and 0.5 g/cm³. Plans were calculated independently using AAA and AXB with identical beam geometries, SBRT prescription and constraints. Dose–volume histograms (DVHs), target dose coverage, maximum dose, and MU/Gy were compared.
Results
For field sizes <3×3 cm², AAA significantly underestimates FOFs relative to Monte Carlo, with discrepancies approaching 9%. AXB shows improved performance, maintaining agreement within ±5% for most small fields and near zero deviation for fields ≥1.2×1.2 cm². For the SBRT plans, AXB predicts reduced target dose coverage, sharper DVH gradients, and consistently higher MU/Gy compared with AAA. These differences become more pronounced for smaller target (1.5 cm diameter) and for lower lung densities (0.2–0.3 g/cm³), where electron disequilibrium effects dominate. AAA’s simplified dose kernel scaling leads to overestimated target dose and underestimated MU, potentially making clinically relevant under dosage in low-density lung.
Conclusion
AAA overestimates target dose and underestimates MU for small-field lung SBRT, particularly in low-density lung, due to its limited modeling of electron disequilibrium. AXB provides closer agreement with MC and more accurate dose predictions for highly heterogeneous small-field conditions and should be preferred for small‑field lung SBRT planning.