Geometry and orientation influences beam hardening in a new high-throughput bone imaging cassette system for microCT

Primary and metastatically derived bone cancers can significantly change bone structure and composition. Visualization and quantification of changes in bone morphology and bone mineral density using high-resolution microCT is a key diagnostic measurement of disease progression and treatment efficacy. The scan quality needed to measure small changes within internal bone structures of mice requires a smaller field of view and longer imaging times across high quantities of samples. Bones are also individually scanned to ensure data quality is not corrupted by beam hardening, a well-known microCT artifact that can reduce the accuracy of calibrated density measurements. In this study, we found that the geometry and corresponding orientation of multiple bones within the field of view are crucial to ensure accurate quantification of cortical bone. By combining the fast-scan rates of the Quantum GX3 microCT (Revvity, Inc.) and a high-capacity (18) bone holder, we report a high-resolution/high-throughput (6 µm, 40 s/bone) microCT imaging methodology, without sacrificing statistical significance in quantification of cortical bone due to beam hardening. To confirm the optimal configuration, healthy bones harvested from age-matched mice with similar mean HU values for cortical bone were arranged in different patterns (2/3/6 bones) in a customized, 3D printed bone holder using low a density plastic scaffold. As a reference, the same bones were scanned individually to compare the effect of beam hardening attenuation in cortical bone (defined as >3000 HU). The mean HU values for each bone's cortical tissue were used to assess the effect of the geometry of one (control), two, three and six bones. Each configuration was scanned in the same field of view (18 mm), time (4 mins) and resolution (6 µm) at soft (70kV, 160 µA,1.0 mm Al filter) and hard (85 kV, 140 µA, 0.5/0.06 mm Al/Cu) x-ray conditions. Calculated HU values (Analyze 15) and the dimensions of the stage and bore were used to design and 3D-print three interlocking cassettes, each with a capacity of six bones for a total of 18 per image. In both hard and soft x-ray images measurable differences in mean HU values were observed, confirming the influence of beam hardening is directly influenced by the geometry of the bones in the holder; however, when compared to the control, a maximum of 4% or 5% difference was observed for any bone in hard and soft x-ray conditions, respectively. Our results show potential for high-throughput and accurate quantification of high-resolution bone scans in the Quantum GX3 microCT. Future work will focus on refining the cassette's design to improve sample loading, evaluating time savings, increasing flexibility in sample type, and providing seamless software integration with AI-enabled segmentation tools.