55. Using experimental designs for validating a unique beam model for four accelerators. (December 2017)

• Record Type: Journal Article
• Authors: Dufreneix, S.; Legrand, C.; Di Bartolo, C.; Bremaud, M.; Mesgouez, J.; Tiplica, T.; Autret, D.
• Is Part Of: Physica medica. Volume 44(2017)Supplement 1
• Page Start: 27
• Publication Date: 2017-12
• Subjects: Medical physics -- Periodicals; Biophysics -- Periodicals; Biophysics -- Periodicals; Imagerie médicale -- Périodiques; Radiothérapie -- Périodiques; Rayons X -- Sécurité -- Mesures -- Périodiques; Physique -- Périodiques; Médecine -- Périodiques; 610.153
• DOI: 10.1016/j.ejmp.2017.10.121
• Languages: English
• ISSNs: 1120-1797

Abstract

Abstract : Introduction: While commissioning a linear accelerator, depth dose, radial and diagonal profiles as well as output factors have to be measured in multiple setups before being entered into the treatment planning system (TPS). The beam model then needs to be validated by comparing computed and measured doses. The objective of this study is to use the same measured dataset for the beam model of four different accelerators and to validate the beam models using experimental designs. Coming from the industry, this method optimizes the number of tests to be performed by methodologically varying several parameters at the same time. Methods: Raw data were measured on a Novalis TrueBeam STx and were entered in the TPS for the Novalis as well as for a TrueBeam Tx, a Trilogy and a C-linac 2300 iX. All accelerators were matched by Varian for the 6 and 16 MV. A Taguchi matrix L36 was selected to validate the AAA algorithm. The use of experimental designs enables testing the following factors in 72 tests: energy, MLC, depth, X, Y1 and Y2 field size and wedge. Results: The mean deviation between calculated and measured doses was 0.1 ± 0.5%. The maximum deviation was 2.4%, below the 3% tolerance recommended in the IAEA TecDoc 1580. Considering the accelerators separately, the mean deviation was −0.1% for the Novalis, 0.3% for the TrueBeam, 0.2% for the Trilogy and 0.1% for the C-linac. None of the investigated parameter was found clinically significant. Results could be correlated to the accelerator type by statistical tests (Novalis and TrueBeam on one side, Trilogy and C-linac on the other). Conclusions: Only one dataset of measurements can be used for several accelerators matched by the vendor. The study conducted could conclude that the four accelerators were dosimetrically equivalent for conformational treatments excluding modulated beams. The use of experimental designs led to a validation of the beam models in 72 comparison points which is time efficient during commissioning.