Reproducibility and error sources of μ-MRI-based trabecular bone structural parameters of the distal radius and tibia

Abstract

The mechanical competence of trabecular bone is significantly determined, next to material density, by its three-dimensional (3D) structure. Recent advances in micromagnetic resonance imaging (μ-MRI) acquisition and processing techniques allow the 3D trabecular structure to be analyzed in vivo at peripheral sites such as the distal radius and tibia. The practicality of μ-MRI-based noninvasive virtual bone biopsy (VBB) for longitudinal studies of patients hinges on the reproducibility of the derived structural parameters, which largely determine the size of the effect that can be detected at a given power and significance level. In this paper, the reproducibility of μ-MRI-derived trabecular bone structure measures was examined by performing repeat studies in six healthy subjects in whom the distal aspects of the radius and tibia were scanned with a 3D spin-echo sequence at 137 × 137 × 410 μm 3 voxel size. Bone volume fraction (BV/TV) and digital topological analysis (DTA) structural parameters including the topological bone surface-to-curve ratio (SCR) and topological erosion index (TEI) were evaluated after subjecting the raw images to a cascade of processing steps. The average coefficient of variation was 4–7% and was comparable for the two anatomic sites and for all parameters measured. The reliability expressed in terms of the intraclass correlation coefficient ranged from 0.95 to 0.97 in the radius and 0.68 to 0.92 in the tibia. Error analysis based on simulations suggests involuntary patient motion, primarily rotation, to be the chief source of imprecision, followed by failure to accurately match the analysis volumes in repeat studies.