Fabrication of 3D printed hollow spheres for quality control and feasibility for use with xSPECT bone

Quantitative accuracy and constancy of Siemens xSPECT Bone quantitative reconstruction algorithm (xBone) can be monitored using activity-filled hollow spheres, which could be 3D printed (3DP) to increase accessibility to phantoms. One concern is that 3D prints can have air gaps in the walls which may pose issues for attenuation correction and xBone tissue zone mapping. This study assessed the feasibility of using 3DP spheres (3DP-S) with materials PLA, PETG and Resin as substitutes for commercial hollow spheres (C-S). Phantom preparation and acquisition parameters were based on the white paper. A Jaszczak phantom was fitted with six ^99mTc- and contrast-filled 3DP-S. SPECT/CT acquisitions were performed on the Siemens Intevo T6 and reconstructed with xBone. Regions-of-interest for activity concentration measurements were drawn to the internal diameter of the spheres. PLA and PETG printed via filament freeform fabrication resulted in minute air gaps, mainly at steep overhang however this did not impact xBone zone maps. Activity concentration recovery of the 3DP-S were within ± 5% of C-S when sufficient projection angles are used (P < 0.05 per equivalence two one-sided test). Spheres printed via masked stereolithography experienced minor resin pooling and increased wall thickness-the smallest sphere was not usable. Resin printing achieved the best watertightness and transparency. PLA and PETG were most affordable but construction was labour intensive. PLA performed best overall in print reproducibility and quantitative accuracy. Similarly printed hollow spheres can be used for quality control of xBone accuracy where C-S are not available. While 3D printing increases accessibility to phantoms, close oversight is required of printing conditions.