S2409
Physics - Quality assurance and auditing
ESTRO 2026
methods, as well as the applied evaluation criteria, may differ in their sensitivity to various types of delivery errors, and the overall consistency of their results remains insufficiently investigated.This study aimed to compare the performance of EPID-based and logfile-based PSQA methods in detecting delivery errors in VMAT plans for prostate treatment. Material/Methods: A VMAT prostate treatment plan was prepared for an anthropomorphic phantom. Based on the original plan, delivery errors were introduced to assess their detectability. The simulated errors included MLC misalignments, collimator rotations (0.5-3°), number of monitor unit (MU) variations (2-10%), and field size modifications (2-3mm). PSQA was performed using the SunCHECK PerFRACTION platform (Sun Nuclear, USA), with pre-treatment (F0) and transit in-vivo dosimetry (FN). In the EPID-based approach, images were compared to predicted with gamma criteria of 3%/2mm (F0), and 3%/3mm (FN), with gamma passing rate (GPR) ≥ 95%. The logfile-based 3D dose calculation was performed on the original planning CT (F0) and on the CBCT (FN). The mean dose values (Dmean) for the target and critical organs were analyzed, with a tolerance of ±3%. Results: The most clinically relevant errors (MLC misalignment errors, MU deviations of 5-10%) were recognized by both EPID- and logfile-based methods, for both F0 and FN. For EPID-based analysis, these errors resulted in GPR ranging 58.9-92.4% (F0) and 60.7-90.9% (FN), with dose differences exceeding the ±3% tolerance for: target volumes and organs at risk (OARs). Differences in response between the methods were observed for minor geometric delivery errors. For MLC position shifts of 0.5 mm: GPR values were 99.8-100% and 97.0- 99.3%; maximum Dmean differences: 4.8% and 6.8% for the OARs, for F0 and FN, respectively. For collimator rotations of 2–3°: GPR: 91.9-99.3% and 94.8- 98.5%; maximum detected differences: 1.15% and 3.3% for the bladder and rectum, for F0 and FN. The effects in number of MU variation (2-3%) were at the edge of detectability according to the acceptance criteria. GPR results were 94.4-100% and 97.3-99.9%, maximum differences ranging 1.8-3.5% and 2.4-4.8%, for F0 vs FN, respectively. Conclusion: EPID- and logfile-based PSQA methods effectively detected clinically relevant delivery errors. However, due to their individual limitations, the combined use of both approaches is recommended to enhance the overall reliability of PSQA. Keywords: PSQA, EPID and logfile-based verification
the strongest predictors of failing plans. Applying all nine complexity metric tolerances in combination successfully identified each of the 53 ‘failing’ plans.
Figure 2. Bar chart illustrating the complexity metric parameters that were the strongest predictors of 'failing plans'. The percentages represent the proportion of ‘failing’ plans that exceeded one or more complexity metric tolerance Conclusion: Leveraging a large, diverse dataset, this study was able to develop and validate an in-house software tool that automates the calculation of complexity metrics and establishes tolerances to accurately identify plans of increased complexity requiring PSQA. This approach has reduced the number of annual pre-treatment PSQA measurements by 80%, enabling early detection of complex outlier plans, and efficiently supporting high throughput VMAT delivery. References: 1. Chiavassa, S., et al., (2019), “Complexity metrics for IMRT and VMAT plans: a review of current literature and applications.” BJR, 92(1102) Keywords: PSQA, complexity metrics, VMAT Digital Poster 5089 Comparison of EPID- and logfile-based PSQA methods in dose delivery error detection: phantom study Marta Kruszyna-Mochalska 1,2 , Barbara Bajon 2 , Kinga Graczyk 2,3 , Krzysztof Matuszewski 2 , Adrianna Wojtalik 2 , Piotr Romanski 2 , Paulina Jasiewicz 2 , Tomasz Piotrowski 1,2 1 Department of Electroradiology, Poznan University of Medical Sciences, Poznan, Poland. 2 Department of Medical Physics, Greater Poland Cancer Centre, Poznan, Poland. 3 Doctoral School, Poznan University of Medical Sciences, Poznan, Poland Purpose/Objective: Patient-specific quality assurance (PSQA) is essential for ensuring the accuracy of radiotherapy dose delivery. Both EPID-based and logfile-based QA approaches enable the detection of potential deviations that may compromise treatment effectiveness or patient safety. However, these PSQA
Made with FlippingBook - Share PDF online