S2176
Physics - Inter-fraction motion management and daily adaptive radiotherapy
ESTRO 2026
models of the two 3D-printed inserts, including (1) bladder, (2) prostate, and (3) rectum.In addition, measurements in a photon-counting CT achieved an HU range of -10 to 150 HU within the phantom. The 3D-printed organs showed a systematic offset of around +50 HU compared to the patient reference data. The adaptive contouring of Varian ETHOS recognized around 75% of the organ volume of the bladder and prostate and 20% of the rectum volume. Conclusion: In this study, a modular anthropomorphic prostate phantom was developed that can replicate human anatomy in terms of shape and image contrast in CT. According to OART the adaptive contouring system identifies the organs inside the inserts well, apart from the rectum. In the future, a fully 3D-printed phantom will be developed, which includes additional surrounding structures such as muscle to improve the phantom’s use for quality assurance in OART. Keywords: 3D-printing, phantom, quality assurance Evaluation of Commercial Deformable Image Registration Algorithms Using Numerical and Physical Phantoms Elena Borderias Villarroel, Farzam Sayah, Marine Stadler, Laure Vieillevigne Medical Physics Department, Univ Toulouse, Oncopole Claudius Regaud, IUCT-Oncopole, Toulouse, France Purpose/Objective: Accurate deformable image registration (DIR) is essential for precise dose accumulation in applications such as adaptive radiotherapy and re-irradiation evaluation. This study aimed to assess the performance of different DIR algorithms implemented in three commercial software programs: AutoContour v2.6.5 (RadFormation), Eclipse v18 and Velocity v4.2 (Varian, Siemens). Material/Methods: Digital Poster 4227 The numerical phantom from AAPM TG-132 (TG132- DP) and the PTW RUBY Adaptive physical phantom (RUBY-Adpt) were used. DIR accuracy was assessed by comparing DIR propagated against manually delineated reference contours. Quantitative evaluation was performed using the Dice Similarity Coefficient (DSC), Mean Distance to Agreement (MDA), and Maximum Hausdorff Distance (HDmax). For each software platform, the impact of the Volume of Interest (VOI) size on DIR performance was investigated by comparing reduced-VOI and full- image-VOI configurations (Figure 1). For Velocity, the Multipass-MP and Extended-EXT (smaller grid resolution and 100 iterations, instead of 30) modes were also compared.
Digital Poster 4194
Developing an anthropomorphic pelvis phantom using 3D-printing for quality assurance in adaptive radiotherapy Maxine Nowak 1,2 , Svenja Schwarz 1,2 , Mona Lifferth 1,2 , Clemens Lang 1,2 , Christina Stengl 1,3 , Armin Runz 1,2 , Peter Häring 1,2 1 Division of Medical Physics in Radiation Oncology, German Cancer Research Center (DKFZ), Heidelberg, Germany. 2 National Center for Radiation Research in Oncology (NCRO), Heidelberg Institute for Radiation Oncology (HIRO), Heidelberg, Germany. 3 Department of Radiation Oncology, Heidelberg University Hospital (UKHD), Heidelberg, Germany Purpose/Objective: Online adaptive radiation therapy (OART) adjusts the radiation plan to the interfractional anatomical changes, which leads to improved dose coverage in tumor tissue and better protection of surrounding tissue. For example, the variable filling volume of the bladder and rectum causes changes in the position and the shape of the prostate. To ensure the quality of OART workflow, new tools are necessary. Therefore, a new method of 3D-printing is used to manufacture a modular prostate phantom for quality assurance (QA) with two interchangeable 3D-printed inserts, which differentiate in bladder fillings. Material/Methods: The phantom depicts a male body in the hip area based on patient data. It consist of a hollow mannequin made from transparent polystyrene plastic as an outer shell, a 3D-printed hip, water with 8.3% potassium chloride as a filling, and one of two 3D-printed organ inserts. One organ insert includes the organs, bladder, rectum, and prostate. They were segmented individually out of CT scans and then combined in two 3D-modeled inserts, insert A with a full bladder and insert B with an empty bladder. For the 3D-printed parts the StratasysTM J850TM Digital Anatomy Printer was used, which allows printing adjustable image contrast like Hounsfield units (HU). Results: The insert change can be performed in a reproducible manner with a maximal deviation of 1.69 mm. (Figure 1).
Figure 1: Assembled modular pelvis phantom and 3D-
Made with FlippingBook - Share PDF online