Material characterization and selection for 3D-printed spine models

he two most popular models used in anatomical training for residents, clinicians, or surgeons are cadavers and sawbones. The former is extremely costly and difficult to attain due to cost, ethical implications, and availability, while the latter is said to not have the same tactile fidelity or mechanical properties as human bone. This study examined the potential use of 3D-printed phantoms to emulate cadaveric, human vertebrae, in hopes of acting as a future use over cadavers. In so doing, we developed 3D-printed MedPhantom®, with the intended use to offer similar tactile feel, mechanical characteristics, and visual appearance as human bone. In order to quantify tactility, a mechanical test was developed where a 5-mm diameter diamond-coated bur spinning at 75,000 RPM swept across the specimens while continuously recording the resultant forces (N) and moments (N-cm), The bur sweep motion is common in orthopedic surgery and neurosurgery. Since most 3D-prints do not offer internal, trabecular structure similar to bone, an algorithm was written to create a stochastic framework of internal mesh to mimic cancellous bone within an STL (stereolithography) file. The ranges of mesh parameters were chosen after several visits with the neurosurgeons participating in the project. In order to quantify structural combinations of wall thickness, gap sizes, and varying cylindrical radii within a print, 1000 RPM compression test with a 5-mm diamond-coated bur was performed with resultant forces (N). Two sample t-test shows statistical significance that samples are not equal to the vertebrae (p < 0.05). Results from the bur sweep test showed 15% Gypsum® powder mixed with 100% Clear® Formlabs resin and 10% Castable® resin mixed with 90% Clear® resin were nearest to human, cadaveric vertebrae, with the difference of force and moment in the x-direction at only 5 N and 7–9 N-cm, respectively. Structural compression results showed that a 2 mm cortical wall, 4 mm or 5 mm gap size between cylinders inside the structure, and 0.25 mm radius of internal cylinders were the best fit parameters to match human vertebrae.