S1648
Physics - Detectors, dose measurement and phantoms
ESTRO 2026
in Germany: a DGMP Working Group survey. Z Med Phys. 2024. PMID: 39414456[2] Shariff M. et al., Multicenter comparison of TBI techniques in Germany. In press, 2025. Keywords: Anthropomorphic Phantom, TBI, OSL Digital Poster Highlight 2232 Commissioning of an EPID-based 3D dosimetry tool using an ESTRO protocol and an independent reference detector Francesco Braglia, Giulia Paolani, Andrea Botti, Ayman El Ouati, Mauro Iori, Roberto Sghedoni, Valeria Trojani, Laura Verzellesi, Elisabetta Cagni Medical Physics Unit, Azienda USL-IRCCS di Reggio Emilia, Reggio Emilia, Italy Purpose/Objective: RadCalc® (RC) software reconstructs 3D dose distributions on patient’s CT dataset by using Electronic Portal Imaging Device (EPID). It operates in two acquisition modes: pre-treatment (in-air) and in- transit (in-vivo). EPID pre-treatment workflow is phantom-less, reduces linac measurement time, and allows direct visualization of dosimetric differences on patient’s CT. RC EPID module, employing back- projection, consists of a Collapsed-Cone Convolution (RC-CCC) dose kernel for independent calculation and a deconvolution model. This study focused on evaluating RC pre-treatment as independent verification tool for plan validation, following an ESTRO commissioning protocol [1] and using Octavius-4D phantom (institutional reference for pre-treatment) for comparison. Material/Methods: The study was conducted on a Varian-TrueBeam® linac equipped with an aSi-1200 EPID. First, RC calculation models were validated on non-clinical test plans following the ESTRO protocol for transit dosimetry [1], here adapted for pre-treatment (Test 0). For each beam energy, phantom tests were performed considering both RC-CCC dose kernel and full EPID-air reconstruction. Water-equivalent slab phantoms simulating homogeneous and inhomogeneous conditions were used. Secondly, two tests were performed on 40 clinical plans (10 IMRT, 30 VMAT). Test 1 compared Octavius-4D measurements (equipped with PTW 1500 matrix) with EPID-air reconstructions on the same Octavius CT (EPIDoct).
energies (Figure 1). For homogeneous configuration, the GPRs range was 84-99%. For inhomogeneous phantom, the lower limit decreased (35-98%); considering AXB plans only, it reached 63%.
For Test 1&2 (Figure 2), moderate, but statistically significant correlations were found between the different GPRs (ρ1=0.43 and ρ2=0.42, respectively).
Test 2 compared the systems in their routine modalities, considering EPID-air
reconstructions on patient’s CT dataset (EPIDpat). All tests were evaluated in terms of Gamma Passing Rates (GPRs) against TPS [2]. Results: In Test 0, a similar trend was assessed for all beam
Made with FlippingBook - Share PDF online