Using mathematical models and advanced control systems techniques to enhance neuroprosthesis function.

Systems that use electrical stimulation to activate paralyzed muscles, called "neuroprostheses", have restored important functional capabilities to many people with neurologic disorders such as spinal cord injury or stroke. However, the clinical benefits derived from neuroprostheses have been limited by the quality of control of posture and movement that has been achieved. Over the past few decades, engineers have used mathematical models and control systems technology to develop functional neuromuscular stimulation (FNS) control systems that show promise in the laboratory, but these have not yet been incorporated into practical solutions for clinical problems. This article briefly reviews several of the complicating factors in controlling FNS systems and describes the potential roles of biomechanical modeling and advanced control system technology. Three important challenges in FNS control systems research and development are identified: 1) to obtain an improved understanding of the biomechanical system that we are trying to control and how it is controlled by the intact neural system, 2) to develop new control system technology with a particular focus on strategies that mimic those used by biologic systems, and 3) to integrate the knowledge and technologies into useful systems that meet the needs of neuroprosthesis users. The outlook for the future includes many interesting problems; yet more importantly, it includes relevant clinical benefits to be gained through the application of biomechanical models and advanced control systems techniques in neuroprostheses.