ESTRO 38 Abstract book

S363 ESTRO 38

conventional phantoms are constructed of rigid materials that lack MR signal. This lack of signal causes the tumor and surrounding tissue to be indistinguishable in MR imagers. Therefore, IROC-Houston’s conventional phantoms are not adequate for MR/CT workflows and new end-to-end QA head and neck (H&N) and thorax phantoms must be constructed for MRIgRT systems. The main purpose of this study was to design and manufacture a stationary H&N and dynamic thorax anthropomorphic QA phantom that could be used as an end-to-end tool to credential institutions for MRIgRT systems. These phantoms were designed for MRIgRT systems that have a magnetic field ranging from 0.35T to 1.50T and were constructed with MR/CT visible and dosimetrically tissue equivalent materials. An MR conditional pneumatic system was also designed for the MRIgRT thorax phantom to enable lung motion during CT simulation and treatment. With the purpose of these phantoms being used as a remote end-to-end auditing tool for credentialing institutions in NCI-sponsored clinical trials, these phantoms were also evaluated through a reproducibility and miniature feasibility study. The reproducibility study was conducted by irradiating each phantom three times on a Unity MR Linac system (7MV/1.5T). The miniature feasibility study was performed by sending both MRIgRT H&N and thorax phantom to three institutions and were irradiated on either an MRIdian (Co-60/0.35T) or an MRIdian Linac (6MV/0.35T) MRIgRT system. The phantoms were evaluated using EBT3 radiochromic film and TLDs and used the same dose constraints and passing criteria as IROC-Houston’s conventional H&N and thorax phantoms. This lecture will discuss the designs, constructions and treatment evaluations for both IROC-Houston’s MRIgRT H&N and dynamic thorax phantom. SP-0704 Phantoms in particle therapy to verify Monte Carlo dose calculation P. Wohlfahrt 1,2,3 1 Massachusetts General Hospital and Harvard Medical School, Department Of Radiation Oncology, Boston, Usa ; 2 Oncoray – National Center for Radiation Research In Oncology, Faculty of Medicine and University Hospital Carl Gustav Carus - Technische Universität Dresden - Helmholtz-Zentrum Dresden-Rossendorf, Dresden, Germany; 3 Helmholtz-Zentrum Dresden - Rossendorf, Institute Of Radiooncology - Oncoray, Dresden, Germany Abstract text High-conformal treatment techniques, such as intensity- modulated radiotherapy or volumetric arc therapy in photon therapy as well as pencil-beam scanning in particle therapy, have been developed in the last decades to achieve a high tumor coverage while sparing healthy tissue more effectively. These technological achievements already led to an improved clinical outcome. Further sophisticated developments in imaging for tumor detection and treatment planning (e.g., dual-energy computed tomography, quantitative and functional magnetic resonance imaging as well as targeted positron emission tomography), and dose calculation (e.g., robust optimization and Monte Carlo algorithm) recently enter the routine clinical workflow. These continuous improvements in treatment precision and accuracy are associated with challenges for medical physicists in the verification of its proper functionality and assessment of its remaining uncertainty. Simplified phantoms, which not adequately simulate the geometrical complexity and tissue heterogeneity of patients, are often not sufficient to represent realistic clinical scenarios and to demonstrate the benefits of new advanced technologies. The use of anthropomorphic phantoms of known composition are suitable for such an experimental validation but require measurement setups close to the physical limits. Here, the generation and experimental

imaging and dosimetry purposes. Dynamic phantoms require the addition of organ motion correlated with breathing traces to ensure that they are patient mimicking. Digital or ‘computational’ anthropomorphic phantoms are typically created from acquired tomographic images, segmentation of organs, and specification of organ densities / chemical composition before a final registration of the segmented images into a 3D volume. Computer simulations are extremely adaptable; allowing control of anatomical features using physics based algorithms. Organ trajectory, shape and density do not always truly represent the patient in computational simulations as intended. Moreover, instead of representing patient specific characteristics, most available anthropomorphic phantoms are created using population-averaged characteristics. However, due to the versatility of computational phantoms, multiple phantoms have emerged which represent a wider population without the cost of manufacturing. Physical phantoms have the added steps of manufacturing the organs and placing them inside a body securely. A plethora of in-house and commercial dynamic radiotherapy phantoms have been developed for imaging and dose verification. These range from simple homogeneous phantoms, placed on unidirectional moving platforms, to phantoms with 6 degrees of freedom motion with lung and cardiac deformations. Physical phantoms can be used in Institution-specific protocols and can include pre-treatment imaging, dose delivery and pre- treatment verification including local set-up procedures. Trade-offs between reproducibility and the magnitude of organ deformations are inherent in the manufacture of physical dynamic phantoms. Emerging phantoms will need to account for the increasingly diverse methods of external tracking which must be accurately correlated with internal tumour motion. With the advent of 3D printing technology and advanced material development, it is clear that improvements in physical dynamic anthropomorphic 4D phantoms will follow for radiotherapy validation. Similarly, computational phantoms will advance significantly with improved resolution, texture and deformation models. This presentation will summarize the advantages and limitations of each dynamic phantom type and will discuss if future 4D radiotherapy research should be in the digital or physical domain, or both. SP-0703 MR Linac anthropomorphic end-to-end QA phantoms: IROC-Houston’s experience A. Steinmann 1 , D. Followill 2 1 Ohio State University Wexner Medical Center, Radiation Oncology, Columbus- Oh, Usa ; 2 University Of Texas Md Anderson Cancer Center, Outreach Physics, Houston- Tx, Usa Abstract text The National Cancer Institute (NCI) requires participating institutions intending to use IMRT in NCI-sponsored clinical trial to first become credentialed by demonstrating their ability to accurately deliver radiation. The Imaging and Radiation Oncology Core at Houston (IROC-Houston) has developed various site-specific anthropomorphic phantoms that are used in the credentialing process. These conventional phantoms were originally designed as end-to-end QA tests for traditional CT-only radiotherapy workflow. Unlike conventional linear accelerators, magnetic resonance imaging guided radiotherapy MRIgRT systems use CT images to capture electron density information for treatment planning and MR images to verify treatment setup, treatment guidance and online adapted radiotherapy. Using a single phantom to perform an end- to-end QA test means the phantom must be used and visualized in both CT and MR imagers. IROC-Houston’s