Osteoarthritis (OA), a degenerative joint disease, is defined by articular cartilage degeneration with possible bone remodeling, osteophyte formation, ligamentous laxity, synovitis, and periarticular muscle weakness. This burdens the patient with progressively decreasing function from joint pain, instability, and stiffness. Unfortunately, in most cases of OA, there is no effective treatment to slow or reverse the disease progression. Prior work has shown that muscle cross-sectional area, tissue composition, and intramuscular fat are associated with progression of knee osteoarthritis (KOA), suggesting that a quantitative analysis can lead to more effective physical therapy treatment. While otherwise impractical, time-consuming, and repetitive for a trained clinician, we will use a deep learning algorithm for the task segmenting radiographs of muscle, fat, and bone to provide quantitative biomarkers such as tissue volume and intramuscular fat to assist in identification of those at-risk of KOA progression and tailor treatment regiments. In addition to discovering novel biomarkers, this study aims to assess the following metrics: change in muscle volume by muscle group, change in intramuscular fat, inter-muscular tissue volume, and overall body composition. We predict that a decrease in muscle volume will correlate to a worsening of OA disease progression, however, we aim to discover the relationship between decline in a specific muscle group to increased disease burden by correlating with WOMAC pain score and Kellgren and Lawrence osteoarthritis classification. MRI images are provided by the NIH’s Osteoarthritis Initiative (OAI) dataset.