Correlation Between CBCT and Torque-Based Measurements of Bone Density: A Study on Polyurethane Foam Blocks and Bovine Ribs

Purpose: To assess whether (and to what extent) average torque readings evaluating the bone density of polyurethane foam blocks and bovine ribs are correlated with that of gray values (GVs) measured via CBCT scans, both at the site of reading and at adjacent sites where bone is expected to engage the implant threads. Materials and Methods: Average torque readings were collected on blocks and ribs via CBCT scans that were then analyzed to measure the GVs of purposedly designed regions of interest (ROIs). The ROIs were shaped as concentric hollow cylinders centered on the sites of average torque measurements. The relation between average torque and GVs was investigated through correlation analyses. Results: The correlation between average torque and GVs from blocks was strong (r = 0.94, r² = 0.89, P < .001), which was independent of the ROI size. On the bovine ribs, the correlation was weak but significant (r = 0.23, P = .029); however, it become stronger when denser bone (average torque ≥ 7 Ncm) was tested (r = 0.41, P = .008). This result was independent from the distance of the site where the average torque had been read. Loss of average torque–GV interchangeability observed on bovine ribs was likely caused by intrinsic bone characteristics, such as an abundance of bone marrow and nonmineralized tissue, possibly having a confounding effect on GV measurements, especially at lower bone densities. Conclusions: Within the limitations of the study, average torque values were found to estimate physical bone density just as well as GVs from CBCT scans. In low-density bone, the assessment of mineral bone density was also found to possibly be more informative than GVs. Results of this study suggest that the micromotor may be presently regarded as a complementary tool to CBCT assessment of bone density and quality in the clinical setting.