The biomechanics of the knee has been the target of decades of scientific and clinical studies due to its significant role in locomotion. The joint exhibits high rates of injury and pathological conditions, e.g. osteoarthritis, a debilitating disese influencing more than 26 million only in the United States. As in any other musculoskeletal joint, the mechanical response of the joint, responsible for its function during activities of daily livin, is the result of the complex anatomical construction, the mechanical properties of its tissue structures, and the mechanical interactions between these components. Computational modeling has been utilized broadly: in a descriptive fashion, to mine experimental data with the goal of understanding knee function; and in a predictive fashion, to design implants and assess surgical and therapeutic interventions. Nonetheless, high fidelity models, not only representative of the specimen-specific anatomy but also capable of reproducing specimen-specific joint response and specimen-specific tissue mechanical properties, do not exist. In addition, specimen-specific models, addressing differences in genders, ages, and pathologies, are not available. Development of robust and reliable knee joint models is a daunting task. In silico representations should be supported by elaborate mechanical testing at joint and tissue levels not only to build the models but also to establish confidence in them. In this collaboration with National Centers for Biomedical Computing, our goal is to establish a platform, supported by crowd-sourcing and cloud computing, to enable development of high fidelity knee models. Modeling efforts, while generally applicable to any musculoskeletal joint, will target at young, elderly, and osteoarthritic knees of different genders, supported by comprehensive anatomical and mechanical data acquired in a specimen-specific manner at multiple spatial scales. To accomplish this goal, we will characterize the joint kinetic-kinematic response, and the material properties of the joint's substructures. Anatomical reconstruction will be based on high resolution magnetic resonance imaging. For project management and also to allow community input, model development and dissemination efforts will be supported by the collaborative infrastructure provided by SimTk.org of Simbios, NIH Center for Biomedical Computation at Stanford. Finite element representations of the knee joint will be developed, with the analysis conducted by FEBio, finite elements for biomechanics. To give the community the opportunity to conduct simulations, a computation infrastructure will be provided by a gateway to XSEDE, Extreme Science and Engineering Discovery Environment. An advisory board of clinicians and knee modeling experts will routinely confirm the direction of the project. Adoption of open development practices, utilization of freely accessible software, and enabling cloud-based simulations will provide the opportunity for community-based development and testing of the models. Accessibility to experimentally confirmed comprehensive knee models will provide utmost reusability for the exploration of healthy and diseased knee mechanics and for establishing biomechanical management strategies to accommodate knee dysfunction.
Computational models provide the footing to understand normal and pathological mechanics of the knee joint and to guide simulation-based intervention design. If founded on comprehensive mechanical experimentation, i.e., acquired at joint and tissue levels from healthy and osteoarthritic knees of different genders and ages, confidence in model predictions, a necessity for clinical decision-making, can be established. Free and timely dissemination, supported by open development and accessible cloud-based simulation infrastructure, provides the opportunity to realize the scientific and translational value of such models by national and international research groups, clinical teams, and orthopedic industry.
