ESTRO 2024 - Abstract Book

S5868

RTT - Service evaluation, quality assurance and risk management

ESTRO 2024

Material/Methods:

A BB cube phantom (DoseLab WL3 Phantom) including a 5 mm ceramic ball (target) was placed in the Encompass immobilization system used by the HyperArc technique. Three simulation CT scans were acquired with a slice thickness of 1 mm each one. The first one (CTsim_ref, 190 slices) included the phantom but not the Encompass area where should lies the shoulders of a real patient. This scan was used as reference because of no image resampling occurred during the online CBCT-CTsim task. The other scans were acquired mimicking the scanning range used for a real patient: one with 350 slices (CTsim_350) and the other with 450 slices (CTsim_450). CTsim_ref was loaded by the PVA software by keeping its original 1 mm slice thickness, while the CTsim_350 and CTsim_450 scans were resampled to have slice thicknesses of 1.40 and 1.80 mm, respectively. For all CTsim scans, a Winston-Lutz plan (Plan_WL) was planned by placing the isocenter at 8 cm from the target center. The Off-Axis Winston-Lutz Calculator v. 1.4 Excel spreadsheet (Appl Clin Med Phys. 2023 Feb;24(2):e13816) was used to design the Plan_WL. After plan creation, the following procedure was performed: 1) phantom was setup on the linac couch to be irradiated with the Plan_WL calculated on the CTsim_ref. 2) A 1mm slice thickness CBCT scan was acquired for online registration before delivery. No image resampling was performed by the PVA software in this case. 3) The Plan_WL was delivered to the EPID system such that a portal image was acquired for each irradiated field. 4) The portal images collected for each plan delivery were analyzed using the Pylinac v. 2.3.2 software. The ‟ Max 2D CAX-BB distance” was the targeting metric analyzed in this study. It is the maximum deviation between the ball center and radiation center found over the portal images acquired on each delivered plan. 5) The steps 1 to 4 were also