Osteoarthritis (OA) is a degenerative disease of the joints, but treatment options have remained limited largely due to an inability to detect and monitor OA in its earliest stages. Medical imaging is a valuable tool to non-invasively evaluate changes in bone, soft tissues like cartilage, and the synovium in OA onset and progression. However, there is a need for improved methods that can not only detect structural and compositional changes in these tissues, but also changes in joint function early in the course of disease. This work develops and progresses the application of medical imaging tools to assess whole joint health in early OA.

First, I focus on a magnetic resonance imaging (MRI) technique called gagCEST which aims to image glycosaminoglycan depletion, thought to be the earliest degenerative cartilage changes in OA, with high specificity. However, technical challenges have limited its translational potential at clinical field strengths like 3 Tesla. Here, I make improvements to the gagCEST technique to enable rapid, 3D imaging of the whole knee with improved dynamic range and good repeatability. While this improved technique offers rapid and volumetric imaging at ultra-high field strengths, it was unable to detect differences in cartilage composition between a healthy cohort and a population with mild to moderate OA at a clinical field strength of 3 Tesla.

While MRI is a valuable tool to probe structural changes in soft tissues and bone, imaging of bone function remains a challenge. Combining MRI with [¹⁸F]sodium fluoride positron emission tomography (PET) imaging allows for simultaneous assessment of structure and function. I advance the application of PET imaging in OA by comparing quantitative measures of bone vascularization and mineralization in healthy and OA knees. Hybrid PET-MR imaging revealed spatial relationships between bone metabolism, synovial inflammation, and structural changes in bone and cartilage. This technique has promise to help study structure, function, and interactions of multiple joint tissues in OA onset and progression.

I further applied hybrid PET-MR imaging to study knee joint function by quantifying the short-term response of bone and cartilage to exercise. Imaging techniques that can detect altered joint function are important in the study of OA, since altered joint tissue mechanics have been implicated in OA onset and contribute to joint pain and dysfunction. I developed a method to distinguish [¹⁸F]sodium fluoride uptake from two consecutive PET scans in order to measure bone metabolism before and immediately after squatting or resting in healthy and OA knees. I found that PET measures of bone metabolism showed regional differences in the response to exercise and were more sensitive to the exercise than cartilage T2 relaxation times from MRI. There were many regions with focally large increases in [¹⁸F]sodium fluoride uptake after exercise, both in areas with and without structural OA changes, suggesting that this technique has strong potential to detect early joint dysfunction after loading.

Overall, this work represents a shift toward whole-joint imaging to assess spatial relationships between structure and function of multiple joint tissues.