S2395
Physics - Quality assurance and auditing
ESTRO 2026
Conventional QA, typically performed before treatment, cannot verify the correct dose delivery in real-time, creating a critical gap in the patient pathway [2]. On-Treatment Monitoring (OTM) systems are designed to bridge this gap by providing near-live verification of the dose delivered during treatment fractions [1, 2]. This work details the commissioning of MobiusFX, a log-file based OTM solution, at NHS Tayside to establish its clinical performance, data integrity, and error detection sensitivity. Material/Methods: 20 class-solution prostate patient VMAT plans were randomly selected and their plans modified in Python (using pydicom and pylinac) to introduce a total of 26 known delivery errors (MLC position, jaw position, collimator rotation, and delivered MU). The original and error-plans were delivered on TrueBeam (v4.1) linear accelerators and verified by simultaneous irradiation of a patient-specific QA phantom (Delta4+) to establish a benchmark to compare existing QA methods. MobiusFX (v4.0.2) processed the resulting trajectory logs, performing 3D dose reconstruction on the patient CT. System performance was assessed through Receiver Operating Characteristic (ROC) analysis to determine optimal tolerance thresholds for error detection and comparison with the phantom- based results using 3D gamma analysis. Results: MobiusFX demonstrated high sensitivity and accuracy in detecting simulated delivery errors. Specifically, the system successfully identified MLC leaf errors down to 5 mm (Figure 1) and collimator rotations up to 3O, and when compared to Delta4+ measurements, the log-file based dose reconstruction showed excellent agreement, confirming its comparability to existing verification methods using standard clinical gamma criteria.
Conclusion: Plan complexity varies not only between treatment sites, but also within treatment sites. We identified up to 15% outlier plans and border plans within each treatment site. Refining the definition of a class solution using complexity metrics could give an objective tool for determining whether an individual plan is part of the class solution or an outlier that requires further evaluation. References: 1. Kamer JB van de et al. Code of Practice for the Quality Assurance and Control for Volumetric Modulated Arc Therapy 2015:65. doi:https://doi.org/10.25030/ncs-024.2. Chiavassa S et al. A. Complexity metrics for IMRT and VMAT plans: A review of current literature and applications. Br J Radiol. 2019;92(1102). doi:10.1259/bjr.201902703. Götstedt J et al. Development and evaluation of aperture-based complexity metrics using film and EPID measurements of static MLC openings. Med Phys. 2015;42(7):3911-3921. doi:10.1118/1.49217334. Sander J et al. Density-based clustering in spatial databases: The algorithm GDBSCAN and its applications. Data Min Knowl Discov 1998;2:169–94. doi:10.1023/A:1009745219419. Keywords: Complexity Metrics, Class Solutions Digital Poster 4358 The Mobius loop: Closing the gap in on-treatment monitoring in radiotherapy Michael J Taylor 1 , Emma McIntosh 1 , Sankar Pillai 1,2 1 Radiotherapy, Ninewells Hospital & Medical School, Dundee, United Kingdom. 2 School of Science & Engineering, University of Dundee, Dundee, United Kingdom Purpose/Objective: The increasing complexity of modern radiotherapy techniques, such as IMRT and VMAT, yields highly conformal dose distributions but introduces significant potential for delivery uncertainty and error [1].
Figure 1: ROC curves for 1mm (left) and 5mm MLC error (right) for 3%/2mm (blue), 2%/2mm (orange) and 2%/1mm (black) gamma tolerance criteria Conclusion: This project successfully validated MobiusFX as a robust, sensitive OTM solution, capable of detecting clinically significant errors. Its performance provides a necessary real-time layer of QA, but its true potential lies in advanced functionality, such as: automated root cause analysis of detected discrepancies, integration
Made with FlippingBook - Share PDF online