S2405
Physics - Quality assurance and auditing
ESTRO 2026
2021-0034. Pócza, Tamás, Zsuzsánna Zongor, Barbara Melles-Bencsik, Dóra Zita Tatai-Szabó, Tibor Major, and Csilla Pesznyák. “Comparison of Three Film Analysis Softwares Using EBT2 and EBT3 Films in Radiotherapy.” Radiology and Oncology 54, no. 4 (2020): 505–12. https://doi.org/10.2478/raon-2020- 0049. Keywords: film dosimetry Mini-Oral 4926 Peer review audits remain essential for ensuring accuracy in proton therapy Paige A Taylor, Jessica Lowenstein, Paola Alvarez, Stephen F Kry Radiation Physics, UT MD Anderson Cancer Center, Houston, USA Purpose/Objective: Describe modern errors in proton radiation therapy identified by independent peer review audits. Material/Methods: IROC and the Radiation Quality Assurance Lab provide a variety of peer review QA services to proton therapy centers around the world: remote output checks using thermoluminescent dosimeters (TLD), anthropomorphic phantom audits using TLD and film, and on-site dosimetry audits using a variety of independent ion chamber-based equipment. An analysis of recent audit results was performed to identify failures in modern proton therapy treatment planning and delivery systems. Results: Despite historically good results, the reference output audits identified two proton beams outside of the five percent criterion this year. In one case, the institution thought they delivered 3 Gy, but upon being notified of disagreement with the measured TLD dose, further investigation showed that a beam interruption had caused only 2.7 Gy to be delivered. The anthropomorphic phantoms continue to identify errors in range and dose calculations. One recent example showed a 5 mm discrepancy between the measured film profile and the planned dose in a liver phantom. Figure 1 demonstrates a profile through the phantom target, showing agreement between the film measurements and TPS calculations. Upon further investigation, the institution identified a change in the beam path calculation during delivery that differed from simulation; a recalculation of this plan from the daily IGRT data showed improved agreement with the measured results.
each blind dose. Software A uses a Multigaussian model while Software B uses image J and applies a six-order polynomial for curve fitting. Results: Software A achieved lower standard deviation for both blind films (Blind 1: 0.0567 vs 27.61; Blind 2: 0.0637 vs 34.91) and a smaller percent deviation (0.0458% vs - 0.1639%) for Blind 1 compared to Software B. For Blind 2, Software B provided a smaller percent deviation from true value (-0.66% vs 1.5%). Both systems performed within published uncertainty levels for film (2-3%) (Figure 1).
Conclusion: Both film dosimetry software systems provide accurate and reproducible results for the two blind dose tests (Blind 1: 720 cGy; Blind 2: 850 cGy). The standard deviations obtained with Software A were significantly lower than those from Software B, likely due to its triple-channel analysis approach. The framework described by Beveridge et al. enabled an independent and objective pathway for selecting the most suitable software for future external dosimetry audits. The demonstrated robustness of the triple- channel analysis supports the adoption of the Radiochromic.com platform in formal audit protocols. References: Beveridge, Sabeena, Andrew Alves, Mohammad Hussein, et al. “An International Film Dosimetry Intercomparison to Establish a Multi - center Audit Framework.” Medical Physics 51, no. 12 (2024): 9071– 87. https://doi.org/10.1002/mp.17428. Méndez, Ignasi, Juan José Rovira-Escutia, and Bozidar Casar. “A Protocol for Accurate Radiochromic Film Dosimetry Using Radiochromic.Com.” Radiology and Oncology 55, no. 3 (2021): 369–78. https://doi.org/10.2478/raon-
Made with FlippingBook - Share PDF online