Reproducibility of bone micro-architecture measurements in rodents by in vivo micro-computed tomography is maximized with three-dimensional image registration

Abstract

In vivo micro-computed tomography (microCT) is a new method to monitor longitudinal changes of bone micro-architecture. Common animal models of bone diseases are mice and rats, and it is important to know the reproducibility of the bone measurements in order to correctly interpret results. When performing baseline and follow-up acquisitions, variation in the scan region will influence the parameters, and it has yet to be investigated if three-dimensional (3D) registration can improve the reproducibility. Two typical breeds of mice and one typical breed of rats were scanned four times each using microCT and standard bone morphological and density measurements were calculated. Image registration was used to find the overlapping regions between the scans within each series of animal data and only overlapping regions were analyzed for the bone parameters. Reproducibility was determined for each animal both pre- and post-registration. For the rats, results included a bone volume ratio (BV/TV) precision error of 5.46%, cortical thickness (Ct.Th) error of 1.97%, and tissue mineral density (TMD) of 2.00%. For the BL6 mice, precision errors were 3.00% (BV/TV), 0.95% (Ct.Th), and 0.94% (TMD), and for the C3H mice 2.68% (BV/TV), 1.52% (Ct.Th), and 1.72% (TMD). After image registration there was a significant improvement in reproducibility in most parameters for the rats. In general, metric parameters such as bone volume ratio had better reproducibility than the non-metric parameters connectivity density and structure model index. With 3D registration, reproducibility improved the results obtained by the experienced operators in this study. Registration may serve to equalize reproducibility of operators with different skill levels and across laboratories. It also improves efficiency by reducing the amount of hand-contouring required. This reproducibility data will be important for the interpretation of current and future longitudinal microCT studies.