S2331
Physics - Quality assurance and auditing
ESTRO 2026
Keywords: adaptive, online, PSQA
deviations were +0.1%, -0.2% and -0.6% for TPS, RadCalc-CCCS and RadCalc-MC, respectively. Across all plans, mean GPRs were 98.3% (RadCalc-MC) and 96.0% (RadCalc-CCCS), with better agreement for single lesions (see Figure 1). With local normalization, mean GPRs decreased to 96.7% (MC) and 94.0% (CCCS). Differences between CCCS and MC GPRs were statistically significant (p<0.001).
Digital Poster Highlight 684 Are secondary dose calculation algorithms robust enough for QA in small field stereotactic radiosurgery? Edoardo Mastella 1,2 , Joseph S Perrine 3 , Klarisa E Szilagyi 1 , Antonio Stefanelli 4,2 , Alessandro Turra 1 1 Medical Physics Unit, University Hospital of Ferrara, Ferrara, Italy. 2 Department of Translational Medicine, University of Ferrara, Ferrara, Italy. 3 Medical Physics Department, Victoria Hospital, Quatre Bornes, Mauritius. 4 Radiation Oncology Unit, University Hospital of Ferrara, Ferrara, Italy Purpose/Objective: Stereotactic radiosurgery (SRS) requires highly precise dose delivery due to small target volumes and the limited number of fractions. Independent dose calculation systems are essential to ensure treatment accuracy. However, the simplified modeling of secondary algorithms, although efficient for routine verification, may struggle to reproduce the steep dose gradients characteristic of SRS treatments [1]. This study aimed to evaluate the robustness of agreement between our TPS and two secondary dose calculation algorithms for SRS patient-specific quality assurance (PSQA). Material/Methods: RadCalc v7.3.2.0 Monte Carlo (MC) and Collapsed Cone Convolution Superposition (CCCS) algorithms, previously benchmarked against Philips Pinnacle ³ v16.4.3 TPS [2], were evaluated with a focus on small field dosimetry. The accuracy of 6 MV FFF beam modeling (1 × 1–10 × 10 cm ² ) was assessed by comparing measured depth doses, profiles, and output factors (OFs) obtained with a PTW microDiamond detector.A dataset of 26 patients was then analyzed, including 13 single and 13 multiple brain metastases (BMs) treated with multi-target single-isocenter VMAT plans. Gamma passing rates (GPRs) were computed using 2%/2mm criteria (30% dose threshold), with both global and local normalization. GPRs and their correlation with plan complexity were analyzed using the modulation factor, equivalent field size, average leaf gap, and edge metric.End-to-end tests were conducted to assess the accuracy of the entire treatment chain, including both on- and off-axis measurements. Results: For all calculation systems, depth-dose deviations were within ±1% and field-size differences remained <0.4 mm. RadCalc-MC slightly overestimated the penumbra (on average 0.4 mm), while RadCalc-CCCS exhibited the largest deviations, particularly in cross- plane direction (1.5 mm vs 0.5 mm in-plane). Mean OF
Among the analyzed complexity metrics, only MC- based local GPRs moderately correlated with MF (see Figure 2).
Made with FlippingBook - Share PDF online