[Poster] Contrôle Qualité Pré-traitement des Arcs Dynamiques en Faisceau de Photons sans Cône Egalisateur : Comparaison de Détecteur*

*Pre-treatment quality assurance for flattening filter free dynamic arcs : a detector comparison

Maxime Goubert1, François Husson2, Laure Parent1
1 Département d’Ingénierie et de Physique Médicale, Institut Universitaire du Cancer de Toulouse – Oncopole, Toulouse, France

2 Dosisoft S.A., Cachan, France

Presented at SFPM 2017

ABSTRACT
Introduction:
Pre-treatment quality assurance (QA) with flattening filter free (FFF) beams is challenging because of the high dose rate and the required spatial resolution as small field are usually used. Our current clinical practice is to perform pre-treatment QA with a diode array in a rotational phantom. This study investigates the replacement of this process by an electronic portal imager (EPI) measurement without a phantom in place in order to shorten the QA time on the machine.

Methods: Pre-treatment QA was performed with PTW Octavius 4D phantom equipped with SRS1000 diode array and with Varian aS1000 EPI. Diode array measurements were compared with the dose calculated in Octavius phantom with Varian Eclipse 13.7. EPI measurements was converted to water measurement and compared with water calculation performed in Dosisoft EPIBEAM (ThinkQA 1.0.2.2ALPHA). QA results were compared for Varian Truebeam 10 FFF dynamic arcs of 10 liver stereotactic radiotherapy plans and for plans with dose and MLC errors voluntarily introduced during beam delivery.

Results: Clinical plans covered a wide range of PTV size (average = 129.5 cm3, minimum = 9.8  cm3, maximum = 327.5  cm3) and dose prescription (average = 11.8 Gy, minimum = 6 Gy, maximum =18 Gy) resulting in 1866 MU on average (minimun = 806 MU, maximum = 3072).Diode array and portal imager QA process gave both gamma values higher than 97% for 2%-2 mm criteria (10% threshold). Analysis of plans with errors is pending.

Conclusion: EPI QA process for FFF beams is achievable. It is much faster than diode array process as there is no need for a phantom setup and it provides similar results.