S1695
Physics - Detectors, dose measurement and phantoms
ESTRO 2026
Digital Poster 4770
This demonstrated statistical control provided the justification to safelyreduce the patient- specific QA frequencyfromdaily to once every threedays for Cyberknife.For Proteus One, analysis of Cpm values specifically highlighted th e poor performance of superficial SFUD plans, whichshowed a Cpsignificantlybelow 1. In contrast, other treatment sites and techniques (e.g., IMPT) maintained a high Cpm value of 1.52 and an LCL of 95.4%,indicating a well-controlledprocess. Instead, cross- comparisonmeasurementsrevealed that the Matrixx O NE detector was a primarycontributor to the low Gamma Index pass rates. Conclusion: The developedgraphical interface enabledcentralizeda nalysis of extensive PSQA datasets, providing real- time process visualization, earlydeviationdetection, and long- term performance stabilization. Frequency for CyberK nifewasoptimized basedon historical Gamma pass rate s, allowing a safe reduction to one QA checkeverythreedays.For proton therapy,weswitched to single- depthmeasurementbecause ourroot- cause analysisconfirmedthat dose discrepancies occur
AUTOMATED PATIENT-SPECIFIC QA MONITORING FOR CYBERKNIFE AND PROTON THERAPY USING A GRAPHICAL USER INTERFACE Walid Ouakkad, Quentin Lavieville, Baptiste Lhomel, Gaëlle Angellier, Joel Hérault, Marie Vidal Radiotherapy, Centre Antoine Lacassagne, Nice, France Purpose/Objective: A retrospectiveanalysis of Patient- SpecificQuality Assurance (PSQA) resultsaimed to identify the main sources of variability and propose a qualitative and quantitative monitoring frameworkusing an automatedgraphical in terface to enhance the robustness of QA processes as a new workflow. Antoine Lacassagne Center wasamong the firstcenters to implementRayCyberknife(2022) Raysear ch’s first optimizerfor Cyberknife. Our facilityisalsokno wnto bethe first Proteus One (IBA) installation worldwide (2016), providing a unique perspective on multi- platform PSQA optimization. Material/Methods: More than 400 PSQA measurementsperformedbetween 2022and 2024 wereanalyzed for Cyberknife. Dose distributions wereverifiedusing SRS MapCHECK (SunN uclear) within a StereoPHANphantom and evaluatedun der 2%/2mm and 3%/2mm Gamma criteriausing SNC Patient. For Proteus One (IBA),
redconsistently at the mostsuperficialdepth. Keywords: QA-CyberKnife-Protontherapy
Digital Poster Highlight 4778
STereotactic Arrhythmia Radioablation (STAR) using CyberKnife: a phantom study to verify methods to improve delivered dose estimate using 4DCT P Alafogiannis 1,2 , P Dvorak 1,2 , T Veselsky 3 , Y Shpunarska 4 , L H.G. van der Pol 5 , J Cvek 1 , L Knybel 1 1 Department of Oncology, University Hospital and Faculty of Medicine, Ostrava, Czech Republic. 2 Department of Dosimetry and Application of Ionizing Radiation, Faculty of Nuclear Sciences and Physical Engineering at Czech Technical University, Prague, Czech Republic. 3 Department of Radiation Therapy, Military University Hospital, Prague, Czech Republic. 4 Medical Physics Unit, McGill University, Montreal, Canada. 5 Department of Radiotherapy, University Medical Center Utrecht, Utrecht, Netherlands Purpose/Objective: The management of respiratory and cardiac motion is crucial in STAR due to single-fraction 25Gy dose delivery [1]. CyberKnife (CK) standard approach relies on dose optimisation on expiration breath-hold CT, treatment delivery in free-breathing, and evaluation of dose delivery accuracy using dose accumulation from 4D-CT. The evaluation can depend on the accuracy of
PSQA performedbetween2016 and 2024 withMatrixx One (IBA) was cross-
validatedusing PPC05 ionizationchamber and Lynx scintillatordetector, analyzing data for more than 1000 patients acrossvariouslocalizations and optimization te chniques (SFUD: Single Field Uniform Dose, IMPT: IntensityModulatedProton Therapy).MyQAiON(I BA) software independentlyverified dose calculationco nsistency. An in-house Python-basedgraphical user interface wasdeveloped to centralize QA data, automate Gamma Index analysis, and computestatistical processmetrics (Cp ₘ : ProcessC apability Index, LCL: LowerControl Limit, trend evolution). Results: Overallprocess control wasenhanced. This wasquantifi ed by a rise in the CapabilityProcessMean (Cpm)from0.79 to 1.48 for the (3%/2mm/5%) criteria, while theLower Control Limit (L CL)increasedfrom87.8% to 97.8%, confirming a more stable and robust QA process.
Made with FlippingBook - Share PDF online