Experience with an independent patient QA dose calculation system in multi-linac and TPS environment
PO-1362
Abstract
Experience with an independent patient QA dose calculation system in multi-linac and TPS environment
Authors: Witztum|, Alon(1)*[alon.witztum@ucsf.edu];Ziemer|, Benjamin P(1);Hasse|, Katelyn(1);Natsuaki|, Yutaka(1);Valdes|, Gilmer(1);Hirata|, Emily(1);Solberg|, Timothy D(1);Scholey|, Jessica(1);
(1)University of California- San Francisco, Department of Radiation Oncology, San Francisco, USA;
Show Affiliations
Hide Affiliations
Purpose or Objective
To assess the use of Mobius 3D (Varian Medical Systems, Palo Alto, CA) as an independent dose calculation system for patient specific quality assurance (PSQA).
Material and Methods
A total of 800 inversely planned treatment plans delivered on an Elekta Versa HD (n=244), Varian Truebeam (n=293), and Truebeam STx (n=263) and planned in Pinnacle or RayStation between June 2018 and February 2019 were included in this study. Each plan was sent to Mobius3D (M3D) for an independent dose calculation; an ion chamber (IC) measurement inside the Mobius phantom was also performed. The difference in point dose between both the M3D calculation, the IC measurement and the treatment planning system (TPS) was calculated, and the institutional passing threshold of 5% was applied. The passing rate of plans using the M3D calculation and the IC reading was compared.
Results
A total of 723 plans (90.4%) passed using both the M3D calculation and the IC reading (Versa = 74.6%, Truebeam = 96.7%, STx = 98.1%). Table 1 shows distribution by treatment site, with 4 lung plans and 7 head and neck (H&N) plans failing PSQA using an IC point dose measurement. All plans that failed the IC measurement also failed the M3D calculation. In addition to this, 66 plans failed the M3D calculation but not the IC measurement. The average (standard deviation) difference between the IC measurement and the TPS was -0.6% (1.8%) and between M3D and the TPS was -2.0% (2.5%). Data analysis during this work prompted reassessment of the Versa machine to improve the beam model. Excluding this machine from analysis resulted in an average difference of -0.2% (1.4%) for the IC measurement and -1.5% (2.1%) for the M3D measurement.
.JPG)
Conclusion
Use of Mobius3D as an independent calculation system can identify plans that would fail an ion chamber measurement. The higher sensitivity of Mobius3D could be used to identify a subset of plans that require ion chamber measurements, freeing up significant clinical resources by identifying potentially failing plans without physical measurements. Additionally, Mobius3D can be helpful in identifying deficiencies in linac beam models that may otherwise be difficult to diagnose.