Quantitative MRI for Detection of Cartilage Damage

Abstract

In this chapter we revisit the experimental basis supporting the use of magnetic resonance imaging (MRI) to diagnose cartilage degeneration. We include those MRI parameters that are measured in vivo on clinical scanners (Chapter 23). Clinical MRI can detect severe damage with high accuracy (91%), but provides only moderate accuracy (76.7%) in detecting early damage. MRI measurements of cartilage thickness and volume are accurate (<10%), reproducible (2–10%) and highly correlated (Pearson's r = 0.58–0.997) with non-magnetic resonance measurements of thickness and volume. Quantitative biomarkers for cartilage composition show moderate-to-strong correlations with the histology score (Spearman's ρ = 0.31–0.77), and moderate-to-excellent correlation with cartilage composition (r = 0.26–0.99). Although the MRI biomarkers vary significantly between healthy and damaged cartilage (Cohen's d = 0.39–2.20), only a few studies analyzed their diagnostic value using a non-MRI standard of reference (e.g. histology, arthroscopy). Thus, further evidence is needed to support the claim that quantitative MRI biomarkers can provide added value to clinical MRI. We conclude this chapter with examples of our most recent experiments in the validation of diffusion tensor imaging (DTI) parameters as biomarkers for cartilage damage including the ability to detect damage after mechanical injury, and the validation of a clinical DTI protocol.

A robust diffusion tensor model for clinical applications of MRI to cartilage

“…Diffusion tensor imaging (DTI) has emerged as a technique that can assess both cartilage structure and composition . One of its two commonly used indices, the fractional anisotropy (FA), is sensitive to the anisotropy of motion of water molecules along and across the collagen fibers , whereas changes in PG content affect another index, the mean diffusivity (MD) .…”