S2147
Physics - Inter-fraction motion management and daily adaptive radiotherapy
ESTRO 2026
Digital Poster Highlight 2947
Characterization of HU-to-density calibration for kVCT imaging in Adaptive Radiotherapy workflows Amandine Capacci 1,2 , Sabrina Décaillet 1 , Nicolas Perichon 1 , Cédric De Marco 1 , Mireille Conrad 1 , Mohammad Varasteh 1 , Oscar Matzinger 1,3 , Maud Jaccard 1 , Marie Fargier-Voiron 1 1 Clinique Générale-Beaulieu and Clinique de Genolier, Swiss Medical Network, Genolier, Switzerland. 2 Physics Section, EPFL, Lausanne, Switzerland. 3 Department of Health Sciences and Technology, ETH Zürich, Zürich, Switzerland
High-density inserts showed the largest HU deviations across acquisition protocols (up to 367 HU). Fine-mode scans were more consistent than Normal-mode. ANOVA indicated significant role of kV and mode, with FOV size affecting head protocols. Eight baseline calibration groups were defined (Fig.1). 160 mAs and 440 FOV size were retained as reference parameters. Scenario (1) yielded mean dose deviations < 1 % on phantom kVCTs (except two groups); Scenario (2) < 1 % for pelvis, < 0.2 % for head-and-neck, < 0.3 % for lung (Fig.2). Scenario (3) showed inter-machine agreement within 0.2 %. In scenario (4) deviations were < 0.5 % except one head protocol (maximal deviation 1.4 %). Over-time tests showed three groups exceeding HU variation tolerances1. Conclusion: To our knowledge, this is the first study characterizing all kVCT acquisition parameters on HU-to-density calibration on tomotherapy machines. Baseline groups showed limited dosimetric impact (<1%), though variations depend on physical density and protocol selection. Monthly control might be necessary to ensure conversion curves reliability. Further investigation will compare this method with adaptive planning based on synthetic-CT strategy. References: 1. ANSM, 2023, Décision du 28/02/2023 fixant les modalités du contrôle de qualité des installations de radiothérapie externe et de radiochirurgie. Keywords: Tomotherapy, ClearRT, HU-to-density calibration Digital Poster Highlight 2959 Dosimetric study of MRI-Guided Adaptive Radiotherapy over Conventional IGRT in Liver SBRT Yaxin Liu 1 , Orlandini Lucia 1 , Xianliang Wang 2 , Jinyi Lang 1 , Shun Lu 1 , Lintao Li 1 1 radiotherapy center, sichuan cancer hospital, chengdu, China. 2 radiotherapy, sichuan, chengdu, China Purpose/Objective: Liver cancer is the third leading cause of cancer- related death worldwide, and radiotherapy plays a key therapeutic role. This study compares MRI-guided
Purpose/Objective: The 3D kVCT imaging system integrated in
tomotherapy machines, primarily used for daily patient positioning, could also support dose calculation and adaptive replanning. Reliable HU-to- density calibration is essential, yet variations in HU numbers across acquisition parameters, machines and time raise uncertainties. This study aimed to evaluate the stability of HU-to-density curves under different imaging protocols, define baseline conversion curves, and quantify dosimetric impacts of calibration variations on phantom and patient plans. Material/Methods: A multi-energy phantom (Gammex, Sun Nuclear) with inserts from lung to bone density was scanned using all kVCT protocols on two Radixact systems yielding 48 datasets with varying parameters kV, mAs, Fine/Normal mode (larger number of views per rotation in Fine mode) and FOV size. HU values were analyzed descriptively and with multifactorial Analysis of Variance (ANOVA) to identify influencing parameters. Protocols showing similar dependence in HU numbers were grouped to define baseline calibration curves. Dose calculations (TomoHelical plan, RayStation 2024A) were performed to quantify deviations of mean target dose in four scenarios: (1) using an incorrect baseline curve on phantom kVCTs with a plan optimized on the brain-like insert as PTV, (2) using wrong curves on patient kVCTs (pelvis, head- and-neck, lung), (3) inter-machine comparison of baseline curves, and (4) replacing baseline curves with protocol-specific curves for outliers. HU stability of each baseline was evaluated monthly for 3 months. Results:
Made with FlippingBook - Share PDF online