Methodology for computed tomography characterization of commercially available 3D printing materials for use in radiology/radiation oncology

3D printing in medical physics provides opportunities for creating patient‐specific treatment devices and in‐house fabrication of imaging/dosimetry phantoms. This study characterizes several commercial fused deposition 3D printing materials with some containing nonstandard compositions. It is important to explore their similarities to human tissues and other materials encountered in patients. Uniform cylinders with infill from 50 to 100% at six evenly distributed intervals were printed using 13 different filaments. A novel approach rotating infill angle 10(o) between each layer avoids unwanted patterns. Five materials contained high‐Z/metallic components. A clinical CT scanner with a range of tube potentials (70, 80, 100, 120, 140 kVp) was used. Density and average Hounsfield unit (HU) were measured. A commercial GAMMEX phantom mimicking various human tissues provides a comparison. Utility of the lookup tables produced is demonstrated. A methodology for calibrating print materials/parameters for a desired HU is presented. Density and HU were determined for all materials as a function of tube voltage (kVp) and infill percentage. The range of HU (−732.0–10047.4 HU) and physical densities (0.36–3.52 g/cm(3)) encompassed most tissues/materials encountered in radiology/radiotherapy applications with many overlapping those of human tissues. Printing filaments doped with high‐Z materials demonstrated increased attenuation due to the photoelectric effect with decreased kVp, as found in certain endogenous materials (e.g., bone). HU was faithfully reproduced (within one standard deviation) in a 3D‐printed mimic of a commercial anthropomorphic phantom section. Characterization of commercially available 3D print materials facilitates custom object fabrication for use in radiology and radiation oncology, including human tissue and common exogenous implant mimics. This allows for cost reduction and increased flexibility to fabricate novel phantoms or patient‐specific devices imaging and dosimetry purposes. A formalism for calibrating to specific CT scanner, printer, and filament type/batch is presented. Utility is demonstrated by printing a commercial anthropomorphic phantom copy.