ESTRO 2024 - Abstract Book

S3417

Physics - Dose calculation algorithms

ESTRO 2024

The Analytical Anisotropic Algorithm (AAA) and the Acuros XB (dose to water/medium (w/m)) (AXB w/m ) as calculated in the Eclipse™ TPS and two different implementations of the Collapsed Cone Convolution algorithm in the RayStation ® (RS) TPS and the Sun Nuclear pre-treatment solution DoseCHECK™ (DC) were evaluated and compared with MCw/m. Comparisons were performed using dose-volume-histogram (DVH) parameters such as the mean dose, dose to X% of the volume (D X% ) and the volume receiving Y% of the prescribed dose (V Y% ).

Results:

DVH parameters evaluated for the different clinical plans generally agreed within 3% between AXB w/m , AAA and MC w/m , except for two plans for D 98% (PTV). For example, a relatively large difference of 2.4% for the mean dose to the Planning Target Volume, D mean (PTV), was obtained between AAA and MC w for a thorax (esophagus) plan. RS indicated an overestimation for the investigated DVH parameters, whereas DC demonstrated the opposite tendency of underestimation. Data for D mean (PTV) and D 98 (PTV) are summarized in Table 1. Calculated dose distributions in water by RS agreed with measured data, although the tendency of dose underestimation was confirmed by the DC software. Thus, an implementation of an algorithm may generate various sources of uncertainty resulting in different deviations for a specific clinical plan (Figure 1). Table 1: Maximum and mean difference, standard deviation (SD) and 95% confidence interval (CI) for the parameters D mean (PTV) and D 98% (PTV) for the AAA, AXB w/m , RS and DC algorithms. All the algorithms were compared with MC w/m . The confidence interval was calculated using the t-distribution.