MRI Classification of Cartilage Degradation

Osteoarthritis (OA) is a degenerative disease which affects the synovial joints of around 20 million Americans. It is characterized by progressive changes in cartilage structure, leading to fissures, and total erosion of articular cartilage, causing pain and limiting mobility. There is currently no treatment that can effectively slow the progress of the disease or regenerate damaged cartilage, and there is no method available to diagnose the more mild and moderate damage that occurs in the early stages of the disease. Magnetic resonance imaging (MRI) shows great promise in imaging and diagnosing cartilage damage due to its excellent soft tissue contrast and the ability to use specific scan sequences and contrast agents to quantify macromolecular contents. MRI has been widely investigated both in vitro and in vivo, but establishing an accurate diagnostic method is still a major challenge due to large patient variability, even in healthy populations.

This thesis focuses on two innovations in MRI of cartilage: zonal analysis of cartilage MR parameters and the non-invasive determination of the collagen content of cartilage with MRI. We apply these innovations in three systems: in a surface degradation model meant to mimic early stages of OA, a classification algorithm to diagnose the type and severity of induced degradation, and to the classification and prediction of the compressive modulus of cartilage explants cultured with inflammatory cytokines. We found that zonal analysis of cartilage is much more sensitive to degradation than full thickness averages, that collagen content is an important classifier of damaged cartilage, and that using zonal MRI parameters we can predict the compressive modulus of cartilage explants in a high accuracy.

This research furthers our understanding of the MR evaluation of articular cartilage and reveals the necessity of considering zonal-dependent properties of articular cartilage. Furthermore, combining zonal analyses and chemical contents of the cartilage tissue quantified from MRI allows us to better classify damaged cartilage and predict its mechanical properties. Future work includes examining the current strategy with human OA tissues, and shortening MRI scan time to enhance clinic translatability.