Recent developments in computational modelling of the knee.

Objective: Model calculations of knee joint loading range from an assumption of perfectly rigid articular surfaces to more realistic simulations of cartilage and meniscal deformation. Rigid-body musculoskeletal models simulate knee contact mechanics using the 'bed of springs' method from elastic foundation theory whereas finite-element models discretise each structure into a series of interconnected elements and ascribe material properties to each element. This mini-review describes some of the most recent developments in computational modelling of knee contact mechanics and suggests possible avenues for future improvements. Design: Narrative mini-review. Results: Muscle and joint contact forces can be calculated synchronously at a reasonable computational cost (typically a few hours of CPU time) using rigid-body models and elastic foundation theory whereas similar calculations using fully deformable finite-element models can take several days and even weeks. The main computational expense incurred in finite-element musculoskeletal modelling is the solution of a muscle-force optimization problem. Conclusion: Calculation of muscle and joint contact forces within the framework of a finite-element musculoskeletal model remains challenging. Integrating biomechanical data from human motion experiments with fully deformable finite-element models to simulate knee contact mechanics during dynamic activity is an evolving science. Future work should explore the use of efficient methods such as direct collocation to perform muscle-driven dynamic optimization simulations of movement using finite-element musculoskeletal models. Dynamic optimization may be combined with finite-element modelling to enable predictive simulations of movement so that the effects of changes in musculoskeletal anatomy on knee contact mechanics can be studied more systematically.