Spinal Cord Injury (SCI) is a devastating life-altering event that impacts the patient physically, emotionally, socially and economically. The population of individuals living with an SCI in the United States grows by about 18,000 each year adding up to an estimated 288,000 as of 2017. SCI and its repercussions introduce high levels of morbidity and low quality of life in these individuals. After SCI, immobilization due to paralysis is observed in a vast majority of cases making patients dependent on assistive devices to perform motor functions.

Immobilization severely affects bone health in SCI population due to lack of loading on legs. Studies have shown that about 60% loss of bone mineral density can be seen in just 1-3 years post injury. This phenomenon is known as disuse osteoporosis and makes bone highly vulnerable to low energy fractures while performing daily activities of living.

Functional Electrical Stimulation (FES) is a rehabilitation therapy that uses electrical signals to produce contractions in paralyzed muscles. When FES is used to flex and extend the legs during ergometer rowing (FES rowing), studies have shown cardio-respiratory benefits. However, biomechanical analysis showed that the force at the feet was modest compared to ablebodied rowers, indicating that FES rowing may be inadequate to prevent bone loss and promote bone growth.

This thesis presents a musculoskeletal forward dynamics computational simulation of FES rowing to probe the effects of muscle timing, assistive springs, and muscle atrophy on the knee joint load. A musculoskeletal model of a rower was built in OpenSim and forward dynamics tool was used to a perform muscle activation driven simulation. The work presented suggests that reaction forces through the feet are higher when the rower does not rely on assistive springs but uses higher intensity muscle activation to perform the rowing motion. Individual effects of assistive spring forces, muscle activation levels and timing of activation on foot reaction force and knee force were studied. Computational models of rehabilitative therapy may provide a guide to effective therapy for bone health.