ESTRO 2024 - Abstract Book

S4874

Physics - Quality assurance and auditing

ESTRO 2024

2303

Digital Poster

Investigation of the spatial accuracy of a Halcyon linear accelerator using the MultiMet-WL cube

Mathieu Marot, Jasmin Winter, Klaus Mühlmann, Andrea Schwahofer

Klinikum Esslingen, MVZ Radiotherapy and Radiooncology, Esslingen am Neckar, Germany

Purpose/Objective:

A recent study [1], which compared the spatial accuracy of a TrueBeam linear accelerator with Halcyon (both Varian, Palo Alto, CA), showed that the Halcyon has a similar positional accuracy as the TrueBeam (deviation < 1 mm). This work aims to validate these measurements using the MultiMet Winston-Lutz (MMWL) cube (Sun Nuclear, Melbourne, FL) on a Halcyon. In the process of implementing stereotactic treatments at Halcyon in our clinic, a dedicated QA needs to be established

Material/Methods:

The MMWL cube includes 6 ball bearings of 5 mm diameter (labelled as ISO, 1, 2, 3, 4 and 5) and made of tungsten carbide material, which are distributed in the longitudinal and lateral planes to extend the standard WL test to off axis positions up to 7 cm from the isocenter (Figure 1). The QA plans consisted of 10-fields 3D conformal field-in-field plans for different gantry and collimator positions. For each plan, a 2 x 2 cm2 MLC field centered on each ball bearing using 100 MUs of 6 MV-FFF photon beam. The cube was aligned with the Halcyon lasers. A cone beam CT was acquired prior irradiation to perform a rigid registration with the images of the planning CT. RT dose images were obtained via the EPID. Since the MMWL software expects couch rotations, the headers of the two RT images corresponding to couch rotations 90° and 270° were modified in MicroDicom and the images were rotated in Matlab (The MathWorks, Natick, MA) in order to be processed. The MMWL software calculates the 3D spatial deviation for all ball bearings using all RT images with the different gantry and collimator angles to determine the deviations between the measured and the expected ball center position in each direction (x,y and z). Measurements were performed on a weekly basis, so for each ball bearing the mean 3D spatial deviation was determined with one standard deviation.