ESTRO 2022 - Abstract Book

S297

Abstract book

ESTRO 2022

Conclusion This work reports some preliminary clinical evaluation regarding the clinical relevance of AC in T2 1.5T MR Elekta Unity® T2 sequence for pelvic RT. Our results suggest that the trained ART-Net® model could be a viable clinical alternative to the human expert. Dice coefficients (84.70%) are higher than the ones usually reported in the literature and our blinded clinical evaluation yielded slightly higher acceptability for the AI contours than for EC, at a fraction of the time (around 30 seconds as compared to 15min for the expert). Introduction of additional samples at training, especially for structures with a small sample number, and of multi-vendor T2 MR cohort could be an interesting future direction to further enhance the clinical relevance of the obtained results. Funding: The European Union’s Horizon 2020 (No. 880314).

PD-0333 Using MR- image quality test phantom for validating MRI-based synthetic CT images

P. Wang 1 , M. Liu 1 , J. Wu 1 , D. Zhang 1 , B. Tang 1 , J. Xu 1 , X. Xin 1 , L.C. Orlandini 1

1 Sichuan Cancer Hospital & Institute, Radiotherapy Department, Chengdu, China

Purpose or Objective Synthetic computed tomography (syCT) from magnetic resonance imaging (MRI) is used for dose calculation in MR-Linacs workflow and represent to date one of the key challenges of the procedure. This study aims to evaluate the accuracy of dose calculation of MRI-based syCT by using a MR- image quality phantom. Materials and Methods A periodic image quality test phantom composed of four sections for measuring various aspects of MR scanner performance was used for this study. The phantom is visible on CT and MR scanners, includes sectors with different densities and is easily locked in a repeatable position using the indexed holders on both CT and 1.5 MR-Linac used in the clinical routine of our department; it is, therefore, suitable for this study. Phantom CT images were acquired using a Philips Big Bore CT with a 1.5 mm slice thickness. The different parts of the phantom were contoured to have for each an average electronic density (ED) to use in the syCT calculation. Three different target volumes were delineated: target 1 (10.9 cm3) and target 2 (67.1cm3) in homogeneous areas of the phantom, and target 3 (10.1 cm3) encompassing areas with different densities; corresponding reference treatment plans (plan 1, plan 2, plan 3, respectively) were performed using a clinical template with ten beam angles. To assess the dosimetric error made using syCT in the calculation process, the reference plans were recalculated on ideal syCT imaging obtained from the reference CT by forcing the drawn contours to the average ED; successively, the phantom was scanned with MR-Linac and the plan was recalculated further using MRI-based syCT. Target Dose-volume histogram D95 and D98 (dose received by 95 and 98% of the volume) of the CT, syCT and MRI-based syCT plans were compared. Results Target 1, 2, and 3 D95 and D98 (dose received by 95 and 98% of the volume) dose differences for the plan 1, 2, 3, respectively, were reported in Table 2. Targets within homogeneous areas (target 1 and 2) showed dosimetric differences below 1%, while target 3 differed from the reference more than 5%. Comparison of plan 1 DVH for CT and syCT is reported in Figure 1.