Subchondral bone and cartilage thickness from MRI: effects of chemical-shift artifact

Abstract

Magnetic resonance imaging (MRI) is the modality of choice for visualizing and quantifying articular cartilage thickness. However, difficulties persist in MRI of subchondral bone using spoiled gradient-echo (SPGR) and other gradient-echo sequences, primarily due to the effects of chemical-shift artifact. Fat suppression techniques are often used to reduce these artifacts, but they prevent measurement of bone thickness. In this report, we assess the magnitude of chemical-shift effects (phase-cancellation and misregistration artifacts) on subchondral bone and cartilage thickness measurements in human femoral heads using a variety of pulse sequence parameters. Phase-cancellation effects were quantified by comparing measurements from in-phase images (TE=13.5 ms) to out-of-phase images (TE=15.8 ms). We also tested the assumption of the optimal in-phase TE by comparing thickness measures at small variations on TE (13.0, 13.5 and 14.0 ms). Misregistration effects were quantified by comparing measurements from water+fat images (water-only+fat-only images) to the measurements from in-phase (TE=13.5) images. A correction algorithm was developed and applied to the in-phase measurements and then compared to measurements from water+fat images. We also compared thickness measurements at different image resolutions. Results showed that both phase-cancellation artifact and misregistration artifact were significant for bone thickness measurement, but not for cartilage thickness measurement. Using an in-phase TE and correction algorithm for misregistration artifact, the errors in bone thickness relative to water+fat images were non-significant. This information may be useful for developing pulse sequences for optimal imaging of both cartilage and subchondral bone.