Correction of Coil Induced Intensity Inhomogeneities in Magnetic Resonance for Trabecular Bone Analysis

Introduction Osteoporosis is a degenerative disease characterized by bone loss, and it is associated with significant socioeconomic impact [1]. Trabecular bone analysis is important in clinical studies of osteoporosis, and magnetic resonance imaging (MRI) allows for non-invasive 3D quantitative trabecular bone assessment. MRI of osteoporosis-related anatomical locations are generally acquired with a surface coil in order to obtain sufficient sensitivity and resolution for quantification of the trabeculae, however these coils introduces intensity inhomogeneities which affect the analysis of the trabecular bone structure. An analytical computation of such bias fields can be obtained given coil size and positioning, but with complex interactions between typically at least four channels in the surface coils routinely used for MRI of trabecular bone such an estimation may not be sufficiently accurate. Currently, techniques such as low-pass filtering (LPF) are employed for coil correction [2], however such a technique assumes that the mean intensity is constant across the image and is therefore subject to edge artifacts. This work evaluates the applicability of a fully automatic coil correction by nonparametric nonuniform intensity normalization (N3) in the analysis of trabecular bone parameters in comparison to LPF coil correction, both by correlation between coil corrected data acquired with a surface coil to homogeneous data ex vivo and inter-scan reproducibility in vivo.