Abstracts Category : Other

Development and Validation of a Kinematically-Driven Computational Model of the Patellofemoral Joint

by Jonathan A Gustafson

The patellofemoral joint represents one of the most challenging musculoskeletal systems to understand and manage. Disruption in the normal tracking of the patellofemoral joint can lead to elevated stress and microtrauma to the articular cartilage, cascading to the development of osteoarthritis. To develop effective treatment therapies, relationships between altered patellar motion and subsequent changes in articular cartilage loading must be measured. Computational modeling provides joint-specific changes in contact mechanics, but current techniques are limited by force-based assumptions and lack validation. The objective of this work was to develop a subject-specific modeling framework driven by highly accurate knee joint kinematics as a tool to estimate knee joint contact stress in vivo. First, a repeatable knee joint testing system for simultaneous measurement of patellofemoral joint kinematics and joint contact pressures was established. Measurements of patellofemoral and tibiofemoral translations and rotations were highly repeatable with intraclass correlation coefficients greater than 0.98/0.90 and 0.80/0.97, respectively. Measurements of joint contact pressure were repeatable within 5.3% - 6.8%. Second, a unique patellofemoral modeling framework employing the discrete element method combined with accurate knee joint kinematics was developed using two cadaveric knee joint specimens. Model-generated stresses were validated using experimentally measured pressures. The model predicted the experimental data well, with percent error (%) differences in contact stress distribution being less than 13%, validating the models ability to predict the experimental changes in joint contact. Lastly, this validated model was implemented in a group of individuals with patellofemoral osteoarthritis (n=5) and a control group (n=6) during a downhill walking task. The model predicted unique patellofemoral joint stress patterns between the two groups such that individuals with patellofemoral osteoarthritis experienced greater (58%) lateral facet joint contact stress early within the loading phase of the gait cycle compared to the control group (38%). This dissertation has validated and implemented a novel modeling technique driven by highly accurate, subject-specific kinematics to estimate patellofemoral joint contact stress during a downhill walking task. Future use of these models can provide quantitative evidence of the effectiveness of current patellofemoral treatment solutions and allow for the development of improved rehabilitation strategies.