Exhibit Hall | Forum 7
Purpose: Independent 3D dose calculation algorithms are increasingly being used for secondary verification of treatment plans. This methodology provides more useful information than simple 1D techniques and is recommended by AAPM TG-219. However, as noted in TG-219, there is a lack of literature on expected agreement and action levels for 3D secondary verification. The purpose of this study was to share our findings on the use of Sun Nuclear DoseCHECK (DC) for 3D secondary verification of Lung SBRT plans.
Methods: Twenty lung SBRT plans generated with the Varian Eclipse TPS were retrospectively studied. All plans delivered 50Gy in 4 fractions using two coplanar arcs, 6XFFF, and calculated with the AAA dose algorithm. Plans were exported to DoseCHECK, and evaluated using Gamma (3%,1mm,40% threshold), percent difference (%Δ) in PTV mean dose, max dose, and isocenter dose. Complexity metrics including aperture irregularity and NMCSv were calculated and Pearson correlation with the dosimetry metrics was calculated. All plans were recalculated using AcurosXB and the same analysis performed.
Results: Average %Δ between TPS and DC was -2.9%, 0.5% and 0.4% for the PTV mean dose, max dose, and isocenter dose, respectively. Average gamma pass rate was 91.7±9.8%. Gamma pass rate was inversely correlated with NMCSv metric and positively correlated with isocenter average equivalent path length. Plans calculated with AcurosXB demonstrated better agreement with DC, with average %Δ of PTV mean dose, max dose, and isocenter dose of -1.0%, 0.1%, and 0.7%, respectively. Average gamma pass rate was improved at 99.4±1.1%. For AcurosXB, isocenter dose agreement was correlated with aperture irregularity.
Conclusion: Agreement of DoseCHECK independent dose verification was correlated to plan complexity metrics for both TPS algorithms and overall agreement was better for plans calculated with AcurosXB compared to AAA. Therefore, actions levels for expected agreement will depend on the TPS algorithm being used.