ESTRO 36 Abstract Book

S757 ESTRO 36 2017 _______________________________________________________________________________________________

Purpose or Objective Anthropomorphic phantoms are used in a variety of ways in radiation therapy for both research and quality assurance purposes. Most anthropomorphic phantoms are of generalized patients, but 3D printing technology can be used to fabricate patient-realistic phantoms for special QA and verification procedures. Most 3D printers, however, can only print in one or two materials at a time, so true patient heterogeneity is limited. In this study, we examined two different patient specific, 3D printed phantoms created based on the same patient to determine the accuracy of single and multi-material phantoms. Material and Methods The phantoms used in this study were designed from the clinical CT data for a post-mastectomy patient treated at our institution. The CT data was trimmed to remove the patient’s head and arms to preserve anonymity and simplify printing. Phantom 1 was designed by converting the trimmed CT data into a 3D model with a CT threshold of >-500 Hounsfield units (HU). This model was sliced into 2.5-cm-thick sagittal slices and printed one slice at a time. All slices were printed with polylactic acid (PLA) representing all body tissues, but with air cavities and lower density regions like the lungs left open. Sagittal slices were chosen for their superior fit with each other, and minimal material warping relative to axial slices. Phantom 2 was designed by converting the CT data into three separate 3D models with a CT threshold of <-147 HU for air cavities, -147 to 320 HU for soft tissue, and >320 HU for bone. The models were sliced into 1-cm-thick axial slices, and printed. The slices were printed from the soft tissue model using a custom formulated high impact polystyrene (HIPS) with the air and bone models left open. After printing, the open bone model sections were filled with a liquid resin polymer with an equivalent density to bone. The phantoms were evaluated for their materials and overall accuracy to the original patient CT. Blocks of PLA, HIPS, and the bone resin material were all imaged to determine their average HU. The phantoms were also each imaged and registered with each other and the original patient CT to determine the consistency and accuracy of each phantom. Results The materials used and their properties are summarized in Table 1. Phantom 1 was fabricated from PLA, which isn’t particularly tissue equivalent, but did print relatively consistently. The bone resin and HIPS of phantom 2 more accurately reflect tissue heterogeneity, but have more variations in their printed consistency.

Conclusion Two phantoms were created, one with a single material, and a second with two materials (tissue and bone). These two phantoms provide an ability to more closely simulate the patient and provide a means to more accurately measure dose delivered in a patient surrogate. EP-1436 A newly designed water-equivalent bolus technique enables BNCT application to skin tumor. K. Hirose 1 , K. Arai 1 , T. Motoyanagi 1 , T. Harada 1 , R. Shimokomaki 1 , T. Kato 1 , Y. Takai 1 1 Southern TOHOKU BNCT Research Center, Radiation Oncology, Koriyama, Japan Purpose or Objective The accelerator-based boron neutron capture therapy (AB- BNCT) system was developed in order to enable the installation of safe hospital BNCT. An important feature of AB-BNCT system is its capability of delivering great doses to deep-seated tumors under condition in which a beryllium target and neutron-beam-sharping assembly are adjusted for production of epithermal neutron that is applicable for more types of tumor localization.Conversely, AB-BNCT is less suitable for superficial cancers, such as malignant melanoma. In this study, we developed a newly water-equivalent bolus technique that has no production of prompt gamma ray and no influence on complicating dose calculation, and we evaluated the effect of this technique on treatment quality for a case of malignant melanoma patient. Material and Methods A water-equivalent bolus was prepared as follows. Urethane foam was cut down into the size of 3-cm larger than the superficial lesion, infiltrated with distilled water with deaeration, and covered with a thin film. The simulated patient was played by a healthy man and simulated condition was originated from a malignant melanoma patient with the lesion of 3-cm diameter localized in a sole of right foot. The superficial lesion was bordered by a catheter and covered with a water- equivalent bolus. Using treatment planning system SERA, the tumor is depicted as a region surrounded by the catheter with 5-mm thickness, and also skin is depicted as the other region except for tumor with 3-mm thickness from body surface. A water-equivalent bolus was delineated as water. This was placed into air in calculation in condition with no bolus. For comparison with bolus-like effect of a covered collimator, the outline of an imaginary collimator cover was set as a mass of polycarbonate or a

Figure 1 shows registered images of the original patient image (a), phantom 1 (b), and phantom 2 (c). Tissue density was more accurate in phantom 2, despite some small holes not being filled with bone resin.