ESTRO 2025 - Abstract Book

S2626

Physics - Detectors, dose measurement and phantoms

ESTRO 2025

3342

Poster Discussion Is negative pre-treatment QA result truly negative? A simple method to reduce inclinometer uncertainty and improve gamma evaluation for Octavius 4D Marta K. Giżyńska 1 , Rob J. Kooter 1 , Rens A.A. Vingerhoets 1 , Marta A. Senkowska 2 , Anna Zygmuntowicz-Piętka 2 , Erik Slooten 1 1 Radiotherapy Department, Noordwest Ziekenhuisgroep, Alkmaar, Netherlands. 2 Medical Physics Department, Narodowy Instytut Onkologii – Państwowy Instytut Badawczy, Warsaw, Poland Purpose/Objective: Gamma evaluation is widely used to compare measured and planned dose distributions during pre-treatment QA. For the Octavius 4D device (PTW, Germany), the dose distribution measured with a 2D array perpendicular to the beam axis is reconstructed as a 3D dose, based on the gantry angle as recorded by an inclinometer. In the presented work we determined the angle uncertainty and its influence on gamma passing rate (GPR). We also developed a simple method to reduce this uncertainty. Material/Methods: In institution A, gantry angles for 45 plans recorded with Octavius 4D inclinometer were compared to machine angles stored in log files (LOG) and xcc files were adapted accordingly. Gamma evaluations with 2mm/3% global settings were performed for both original and adapted xcc files. In order to establish a fast and transferable correction workflow which does not require LOG analysis, a gantry angle correction function was created based on static gantry measurements. The xcc adaptation workflow based on this correction function was compared to the LOG based method. Furthermore, the effect of gantry angle and inclinometer angle during the Octavius 4D reference run was tested. Long term stability of the angle correction function is currently being investigated. The methodology was applied in another institute to test its transferability. Results: In Institute A, applying inclinometer angle correction based on LOG improved GPR on average by 0.7% [0%, 2.1%] (Figure 1). Two plans could be accepted against 95% GPR threshold only after applying gantry angle correction. Using the simple angle correction function gave similar results to the more sophisticated LOG based corrections with differences of 0% [-0.4%, 0.5%]. The angle correction functions were different for both institutes (Figure 2a), with deviations of the inclinometer readout in range [-0.1°, 1.2°] and [-1.8°, 0.5°] for institute A and B, respectively. With the gamma evaluation distance to agreement parameter set to 2mm, a 1.5° deviation in angle readout can cause gamma > 1 for points lying further than 7.6cm from the phantom center (Figure 2b). The angle correction function shape/offset depend on initial inclinometer/gantry angle positions while performing reference run.