ESTRO 2025 - Abstract Book

S2948

Physics - Image acquisition and processing

ESTRO 2025

Receive coil holders are commonly used in MR-only treatment planning to prevent body surface deformation. However, their use increases imaging workload and compromises coil-to-body distance, potentially affecting image quality. New lightweight receive coils may enable placing the receive coil directly on the patient’s body surface for MR-only planning. Material/Methods: Twelve patients underwent pelvic MRI simulation in a single session using two different setups without table or patient moving: one with a coil holder (Qfix INSIGHT, 'holder' setup) and another with the receive coil (Contour L) placed directly on the patient’s body surface ('contact' setup). Each patient received a T1-VIBE-Dixon sequence for synthetic CT (sCT) generation in both setups, alongside a conventional planning CT, which served as reference. Comparative analysis was performed between the 'contact' and 'holder' setups in terms of body surface deformation, synthetic CT quality, and dosimetric accuracy. Results: Mean net anterior body surface impression was 0.53 mm (range, -0.10 – 1.15 mm) for the 'contact' compared to the 'holder' setup. Geometric similarity of the body surface between the two imaging setups was very high, with a mean Dice similarity coefficient (DSC) of 0.993 (range, 0.987 – 0.997) and a mean surface distance (MSD) of 0.45 mm (range, 0.23 – 0.88 mm), both of which were significantly better than the repositioning accuracy observed in the planning CT (mean DSC 0.968, mean MSD 2.54 mm, p=0.002). The mean absolute HU error between the 'contact' (53.91 HU) and 'holder' (53.21) setups was not significantly different (p=0.239). Dosimetrically, there were no significant differences between the two setups in mean D2ccm and D50 errors for the PTV (D2ccm, 0.79 vs. 0.82 Gy; D50, 0.76 vs. 0.77 Gy) and OARs, except for a smaller D50 error in the 'contact' setup for the cauda equina (0.51 vs. 0.54 Gy, p=0.022) and right hip (0.22 vs. 0.24 Gy, p=0.047). PTV coverage error was also not significantly different between setups (mean, 1.16 pp vs. 1.18 pp, p=0.515). Furthermore, mean gamma passing rates for 3%/3mm and 2%/2mm criteria were not significantly different between the 'contact' and 'holder' setup sCTs (98.67% vs. 98.50%, p=0.083 and 94.24% vs. 94.07%, p=0.202, respectively). Conclusion: The interim analysis of the MR.pelvis study suggests that pelvic MR-only treatment planning without the use of a receive coil holder is feasible. Placing the receive coil directly on the patient had negligible impact on body surface deformation and did not compromise synthetic CT quality or dosimetric accuracy.

Keywords: MRonly, MRI treatment planning, MRI

391

Digital Poster Individually synchronized contrast-enhanced 4DCT simulation for target volume delineation in abdominal SBRT Valeria Faccenda 1 , Denis Panizza 1 , Rita Marina Niespolo 2 , Riccardo Ray Colciago 3 , Giulia Rossano 3 , Lorenzo De Sanctis 3 , Davide Gandola 4 , Davide Ippolito 3 , Stefano Arcangeli 3 , Elena De Ponti 1 1 Medical Physics, Fondazione IRCCS San Gerardo dei Tintori, Monza, Italy. 2 Radiation Oncology, Fondazione IRCCS San Gerardo dei Tintori, Monza, Italy. 3 School of Medicine and Surgery, University of Milan Bicocca, Milan, Italy. 4 Diagnostic Radiology, Fondazione IRCCS San Gerardo dei Tintori, Monza, Italy Purpose/Objective: Stereotactic Body Radiotherapy (SBRT) is emerging as an ablative, non-invasive option for treating liver and pancreatic tumors. Precise target delineation is crucial due to the high conformality of SBRT. While four-dimensional computed tomography (4DCT) is the standard for respiratory motion assessment, it often struggles to clearly visualize abdominal tumors due to poor contrast with surrounding tissues in terms of Hounsfield Units (HU). This