Task-dependent scheduling of feedback gains in animal motor systems: too much, too soon for FES control?

Summary form only given. The idea that certain parts of the mammalian nervous system, such as the cerebellum, might be specialized for setting transmission parameters elsewhere is discussed. Experiences with the control of electrical stimulation in paretic patients, experiences which have shown that transducer systems are still far too primitive to allow this level of sophistication, even in laboratory-bound devices, are described. At best controllers rely on fixed-parameter feedback from one or two sensors to cycle through a simple finite-state control sequence. Kinematic variables are essentially driven open-loop. The author's approach to functional electrical stimulation (FES) control is constrained by transducer technology, and evolves virtually without reference to the features of the animal control systems described.<
>