SU-E-T-519: Experimental Evaluation of Deterministic Acuros XB Radiation Transport Algorithm for Heterogeneity Dose Calculation Using the Radiological Physics Center's Lung Phantom.

Abstract

To evaluate the heterogeneity corrected dose calculations from the Acuros XB (AXB), a novel deterministic dose calculation algorithm based on grid-based Boltzmann transport equation solver (GBBS), for IMRT and VMAT plans. The Radiological Physics Center's lung phantom was used to create clinically equivalent IMRT and VMAT plans (RapidArc) with the Eclipse planning system 10.0 that were delivered using a Varian 23 iX. Absolute doses and relative dose distributions were measured with thermoluminescent dosimeters (TLDs) and radiochromic film. The measured dose distributions were compared with calculated doses from both AXB (11.0.3) and AAA (10.0.24) dose calculation algorithms. The AXB calculated dose-to-water and dose-to-medium were both compared to measurements. Gamma analysis (±7%/4mm, ±5%/3mm, and ±3%/3mm) was used to quantify correspondence between AXB dose distributions and the film measurements. The computation time between AAA and AXB were also evaluated. For TLD point doses, both AAA and AXB heterogeneity corrected dose calculations are within 5% inside the PTV for both IMRT and VMAT plans. The agreements observed between the measured and calculated doses for both AXB dose reporting methods are better than those observed with the AAA algorithm. The gamma analysis showed that the differences between AAA, AXB and film measurement met the RPC ±7%/4 mm criteria. The percent of pixels passing rate for both the AXB dose to medium and AXB dose to water are higher than AAA. The computation time between AAA and AXB are comparable for IMRT plans but AXB is significantly faster (4 times) than AAA for VMAT plans. The AXB implemented in the Eclipse planning system calculates a more accurate heterogeneity corrected dose than the AAA algorithm as compared to measurement in lung and improve the calculation speed for VMAT radiotherapy. Work supported by grants CA10953, CA81647, 2R44CA105806-02, CA016672 (NCI, DHHS).