Grant: Fellowship: Bisby (Prize)

Osteoarthritis (OA) is the most common joint disorder worldwide and disability attributable to knee OA is equal to that of cardiovascular disease and greater than any other medical condition in older adults. The prevalence of knee OA is projected to double by 2020, due in part to increases in obesity and longevity. X-ray diagnosed OA is the most widely used clinical and research measure but it is not strongly associated with knee pain and doesn't show early OA - a critical period for treatment to prevent disability. Conversely MRI is related to knee pain and is able to show early disease and progression but MRI is expensive due to the time to read and interpret images, and is largely subjective. These factors have resulted in small research samples (due to expense) with limited statistical power to show important changes, and a sometimes confusing array of information for physicians. New MRI software has been developed and pilot-tested that provides readings that are semi-automated and objective (quantitative), leading to faster, cheaper, and more accurate diagnosis. However they have not been validated in large datasets - a necessary step before widespread research and clinical use. This research will compare the new MRI measures to conventional MRI measures, testing them for the ability to correlate with a patient's pain and function, identify disease progression, and reduce reading times. Significance: This work will immediately benefit clinical researchers in terms of improved diagnosis at lower costs leading to larger, more powerful studies to assess patient populations with knee OA. This will strengthen disease prevention efforts and lead to earlier clinical interventions to prevent disability. It will benefit treating physicians through development of a single diagnostic score for knee OA (similar to the single bone density score for osteoporosis) leading to earlier and more efficacious treatment.