17 Patient Quality assurance of the new halcyon linear accelerator (Varian)

Introduction Recently commissioned in our centers, the latest generation HALCYON (Varian Medical Systems) 6xFFF (800UM/ Min) Linear Accelerator was installed with the TPS Eclipse V15.5, all the algorithms and dosimetric parameters of the multileaf collimator that have been preconfigured by the manufacturer: IMRT/ VMAT (PO V15.5.1.1), AAA, portal dosimetry, dosimetric Leaf Gap  = 0.010 cm and slide transmittance  = 0.005 (0.5%). The aim of this study is to validate the quality assurance of the generated plans to ensure the precision of the preconfigured algorithms with respect to the machine and to check if we need to adjust the beam modeling. Methods Twenty-one cases (Head u0026 Neck, Prostate, SBRT lung/ liver/ spine, SRS, VMAT breast, etc.) from international dosimetry competitions, were downloaded from medicalaffairs.varian.com. 15 plans were imported with the RCMI technique and 14 VMAT plans. In addition, 10 breast cancer treatment IMRT plans have also been generated. The dosimetry tools used for validation are: portal dosimetry using the V15.5.1.1 Anisotropic Analytical Algorithm (AAA), the Mapcheck2 matrix (Sun Nuclear) and an Arccheck 3D detector (Sun Nuclear). Results For the portal dosimetry, we obtain on average for the IMRT plans a global gamma (3%, 3 mm) = 99.88 ± 0.19% and local gamma (3%, 3 mm, threshold 10%) = 99.06 ± 0.84%. For the VMAT plans, a mean global gamma (3%, 3 mm) = 99.68 ± 0.39% and a mean local gamma (3%, 3 mm, 10% threshold) = 98.08 ± 1.53% are found. For the 10 IMRT breast plans, the results between the portal dosimetry and the mapcheck2 are close and correspond respectively to a local mean gamma (3%, 3 mm, 10% threshold) = 99.9 ± 0.28% and 99.56 ± 0.65%. Other analyzes are currently being carried out and in particular with the Arccheck 3D detector for VMAT plans. Conclusions The first results show a very good agreement between measurements and calculations for the different controlled plans. The preconfigured algorithms show great robustness when used in clinical routine.