(Wednesday, 7/17/2019) 4:30 PM - 6:00 PM
Room: 304ABC
Purpose: Anthropomorphic phantoms mimicking realistic tissue radiographical properties are important in therapy and imaging practices. Conventional anthropomorphic phantoms are limited in option and expensive. 3D printing is a viable technology for fabricating custom phantoms, but a process to print and assemble anatomically and radiographically realistic phantoms has not been demonstrated. In this study for non-human-primates, we developed and characterized a phantom from CT images using 3D-printed materials that closely mimic the x-ray attenuation properties of soft tissues and bone.
Methods: CT images were acquired for a 3kg male rhesus monkey and segmented by Hounsfield Units (HU) into bone, lung, and soft tissues, with further segmentation for printing and assembly. The phantom features inserts for TLD and film dosimetry and a honeycomb structure within the lungs for accurate density. The Formlabs Form2 printer was used, with Flexible Resin for soft tissue and lung and Ceramic Resin for bone. A treatment plan was created to deliver 8 Gy to a 10mm target in the phantom chest in one full arc, using real tissue, lung, and bone material properties for calculation.
Results: The average CT measurement for soft tissue differed by 26.9 HU between the monkey CT and the phantom (35 HU and 8.1 HU, respectively) for ROIs in the abdomen. The average measurement differed by 19.7 HU for lung (-765.9 HU and -746.2 HU), and 4 HU for bone (902 HU and 906 HU). On the HU scale, this represents differences of 1.3%, 1.0%, and 0.2% for soft tissue, lung, and bone, respectively. Preliminary measurements show <1Gy difference between the delivered and calculated mean target dose.
Conclusion: A CT-specific 3D-printed monkey phantom was developed that closely mimics the CT values for tissue and bone, providing a useful preclinical research tool and establishing a straightforward method for creating more anatomically accurate custom phantoms.
Funding Support, Disclosures, and Conflict of Interest: This work was supported by the Centers for Medical Countermeasures against Radiation (NIH U19AI67798 and U19AI067769) and NIH R21EB025269.
Keywords
Not Applicable / None Entered.
Taxonomy
Not Applicable / None Entered.
Contact Email
