Current treatments for anterior cruciate ligament injury are largely successful at restoring stability to the knee and enabling patients to return to functional activities. However, the long-term outcomes are suboptimal, with high percentages of patients showing signs of early onset osteoarthritis (OA) 10 years post-injury regardless of treatment. In vitro and animal studies indicate that alterations in cartilage loading patterns following injury disrupt tissue homeostasis and are likely a key factor in the initiation of OA. Dynamic imaging studies indicate altered kinematics, and thus loading patterns, are present in ACL deficient and reconstructed knees during functional movements. Thus, improvements in both conservative rehabilitation and surgical reconstruction treatments for ACL injury are necessary to restore cartilage loading and preserve the long-term health of the knee.

Musculoskeletal computer simulation provides opportunity to gain insight into the how modification in treatments effect knee mechanics during movement. However, existing simulation frameworks either focus on resolving detailed knee mechanics with finite element models, or the muscle forces necessary to generate a measured movement using musculoskeletal models. Few simulation frameworks are capable of simultaneously resolving whole-body and joint scale dynamics. Traditionally, despite the substantial uncertainty in model parameters and the objective used to resolved muscle redundancy, musculoskeletal simulations have been performed in a deterministic manner. This dissertation introduces the Concurrent Optimization of Muscle Activations and Kinematics (COMAK) framework to predict muscle forces and knee joint mechanics during movement in a probabilistic manner.

The COMAK framework was then applied to investigate conservative and surgical treatments for ACL injury. The framework predictions indicate that wear patterns in ACL and menisci deficient knees correspond with regions that experience increased contact pressure during walking. To better inform surgical practices, the framework was used to assess the influence of controllable factors (graft stiffness, reference strain, and tunnel locations) on predicted knee mechanics during walking. For conservative treatment, model predictions indicate that cartilage loading patterns during walking cannot be restored solely through altered muscle coordination strategies.