S2343
Physics - Quality assurance and auditing
ESTRO 2026
With Locally Advanced Esophageal Cancer (ARTDECO Study). JCO. 2021;39(25):2816-2824. doi:10.1200/JCO.20.03697 Keywords: quality assurance, clinical trials, OAR
Digital Poster 1292
Can the first choice be adequate? Assessing of gamma analysis criteria for Ethos Systems. Joanna Iwanowska-Hanke, Paulina Wesolowska, Marta Fillmann, Sandra Maluszczak, Bartlomiej Sadowski, Mikolaj Tarchalski, Agnieszka Walewska Mdical Physics Department, Maria Sklodowska-Curie National Research Institute of Oncology – National Research Institute, Warsaw, Poland Purpose/Objective: The aim of this study was to determine clinic-specific, technique-specific (on-line adaptive radiotherapy using ETHOS system), and measurement method-specific (ArcCHECK device) tolerance (TLCS) and action limits (ALCS). It was also to assess whether the parameters of gamma analysis and its passing rate (%GP) adopted at the stage of implementing adaptive radiotherapy for clinical use should not be changed. Material/Methods: In our hospital the Ethos adaptive radiotherapy was implemented, allowing real-time plan adaptation based on patient anatomy. Each clinical plan was verified dosimetrically before treatment using ArcCHECK (SunNuclear) regardless of the Mobius3D® (Varian) calculations Measured and calculated in Ethos system dose distributions were compared using 2D global gamma analysis and initial acceptance level was set to gamma passing rate %GP ≥ 95% (3%/3 mm) with 10% threshold To determine clinic-specific tolerance (TLCS) and action limits (ALCS) 378 plans (279 IMRT, 99 VMAT, doses per fraction 100–550 cGy), were subjected to gamma analysis with criteria 3%/3 mm, 3%/2 mm, 2%/2 mm and 10% dose threshold. According to AAPM TG 218 [1] procedure local TLCS and ALCS limits were determined. Results: Measured and calculated dose distributions were compared using global 2D gamma analysis (3%/3 mm, 3%/2 mm, 2%/2 mm; dose threshold 10%). For each gamma analysis parameters mean value, standard deviation, and TLCS/ALCS were calculated separately for IMRT and VMAT plans (Table 1).
Conclusion: Our results show that the tool provides valuable visual guidance by accurately identifying regions within OAR contours that require review. Future work will include validation on larger datasets and user testing by QA clinicians to assess its practicality and robustness. Nonetheless, the statistically significant results observed indicate that this approach could substantially enhance the efficiency of contour QA in clinical trials. References: 1. De Biase A, et al. Clinical adoption of deep learning target auto-segmentation for radiation therapy: challenges, clinical risks, and mitigation strategies. BJR Artif Intell. 2024;1(1):ubae015. doi:10.1093/bjrai/ubae0152. Kingma DP, Welling M. Auto-Encoding Variational Bayes. arXiv. Preprint posted online December 10, 2022. doi:10.48550/arXiv.1312.61143. Welch M, et al. Computed Tomography Images from Large Head and Neck Cohort (RADCURE). Published online February 2024. doi:https://doi.org/10.7937/J47W-NM114. Hulshof MCCM, et al. Randomized Study on Dose Escalation in Definitive Chemoradiation for Patients
Made with FlippingBook - Share PDF online