ESTRO 2021 Abstract Book

S1316

ESTRO 2021

per case) with a single isocenter. Target diameter ranged from 4 mm to 33 mm. 6 MV photon beams from a TrueBeam linac equipped with a Millennium 120 MLC were used. Plans were computed with the Acuros XB algorithm v. 16.1.0 with a grid size of 1 mm. The preconfigured M3D beam model was used, with the exception of the output factor for the 1×1 cm 2 field size, which was changed to match our measurements. The dynamic leaf gap (DLG) correction factor was used in M3D was set to 0. The Eclipse SRS plans were recomputed by the M3D software using a grid size of 1 mm. The Eclipse SRS plans were also simulated with the PRIMO v. 0.3.64.1800 Monte Carlo software (www.primoproject.net), using the phase-space files provided by Varian for 6 MV photon beams of a TrueBeam. The accuracy of PRIMO calculations was investigated using EBT3 radiochromic film. Three SRS plans treating multiple targets of 5, 8 and 10 mm on a polystyrene slab phantom were designed in Eclipse and simulated with PRIMO. The dose distributions at the coronal plane containing the targets were compared with the respective film dose planes, using the radiochromic.com software. The MD3 plans were compared with PRIMO-based plans. 3D global (G) gamma analysis (M3D vs. PRIMO) was computed for all targets (83).Action level (AL) (AL= 100% - 1.96 SD) for the gamma passing rate (GPR) metric was calculated. Results The accuracy of PRIMO was estimated in 5% (Global)/1 mm, while the uncertainty estimated for the film dosimetry method was of 4% ( k =2%). The statistical uncertainty parameter of the simulations as reported by PRIMO was < 2% ( k =2) for all the plans. The agreement between M3D and PRIMO resulted in GPRs of 96.5% ± 6.4% (AL: 84.0%), 95.8%± 7.2% (AL: 81.7%) and 98.4% ± 3.5% (AL: 91.6%) for 2%(G)/2 mm, 3%(G)/1 mm and 5%(G)/1 mm criteria, respectively. Conclusion The highest action level of 91.6% was obtained for the 5% (Global)/1 mm criteria. A minimum AL of 90% is considered as adequate in this work. Therefore, the Mobius 3D software v. 3.1, as configured in this work, could be used for SRS patient-specific QA using a 5% (Global)/1 mm criteria at least. Purpose or Objective The Agility MLC (Elekta) includes 80 leaf pairs and is designed to deliver highly modulated techniques. The Monaco treatment planning system (TPS) incorporates an adjustable MLC model, in order to calculate the dose plan with high preciseness. The Monte Carlo algorithm included in the Monaco TPS is based in three different components: virtual source model (VSM), transmission probability filters (TPFs) and x ray voxel Monte Carlo. After Elekta provides the user with the Monte Carlo beam model based on the measurements, the MLC parameters can be fine tuned to asses the behavior of the Agility collimator installed. This optimization has to be done adjusting the different TPF parameters. To create a rigorous systematic process, the vendor provides a guide and a set of test fields. Both in the vendor's guide and the works published, this optimization is done using detectors with poor resolution capability. Sometimes it is done with EBT3 films. Nevertheless, the uncertainties associated with the EBT3 readout procedure are sometimes larger than the TPF parameter adjusted. The purpose of this study is to perform an optimization of the TPF parameters in the Monaco TPS, using the measurements provided by an Electronic Portal Imaging Device (EPID). Materials and Methods The Agility head is mounted on a Versa HD linac (Elekta) and all the measurements are performed in the EPID (IviewGT) attached to the machine. The detector is calibrated in absolute dose in water, and it has been tested against other devices (ion chambers and arrays). The procedure suggested by the vendor includes a set of predefined fields called the ExpressQA package. This fields are specifically designed to evaluate different TPF paramaters with each one. Evaluating the dose obtained measuring this fields against the calculated dose, allows the user to fine tune the parameters to achieve better agreement between both. This comparison is performed with the gamma criterion, using 2 % global dose, 1mm distance to agreement and 10 % of the maximum dose as threshold. The plans were calculated in Monaco with a 1mm grid size and a statistical uncertainty of 0.2 % per control point, to reduce significantly the noise contribution. The EPID has a resolution of 0.4 mm at 60 cm from the isocenter (160 cm source to image detector), that is equivalent to approximately 0.2 mm at the isocenter. There are many TPF parameters that can be adjusted. In this work we only studied the Leaf Tip Leakage, the Leaf Offset, the Interleaf Leakage and the MLC transmission. Besides, all of them were adjusted for a 6 MV beam and a 6 MV FFF beam. Results The results for both energies are displayed on the tables. PO-1593 Optimizing Agility MLC model in the Monaco treatment planning system F. San Miguel 1 , P. Chamorro 1 1 Hospital Central de la Defensa 'Gómez Ulla', Medical Physics, Madrid, Spain