A state controller for closed-loop functional electrical stimulation regulated by natural sensory feedback
Year: 1996
Type: dissertation
Abstract: Functional electrical stimulation (FES) is used in clinical systems to return function to patients with extremities paralyzed by spinal cord injury or stroke.Stimulation pulses are applied through surface or implantable electrodes to nerve or muscle tissue below the level of the lesion, resulting in muscle twitch or sustained contraction in paralyzed muscles.Coordinated stimulation of multiple muscles may be required to return useful function to an entire extremity, and FES is currently being implemented to produce standing and walking in paraplegic patients and pinch grip in quadriplegic patients.Closed-loop FES systems may improve the efficacy of clinical systems by including feedback from artificial sensors external to the body or, in a new approach, from natural physiological sensors that remain viable in affected extremities.In this thesis, a state controller for FES was developed to demonstrate the viability of implementing natural sensory signals from nerve cuffs in the periphery as feedback for closedloop FES systems.In a first study, recording nerve cuff electrodes were implanted on two of the Median, Ulnar, andlor Radial nerves in the forelimb of fourteen cats.The functional condition of the nerves and recording devices was periodically evaluated by monitoring evoked compound action potentials (CAPS) and device impedances under anesthesia.The study demonstrated that implanted nerve cuffs are safe and that stable neural recordings are possible for extended periods of at least six months in freely moving animals.In a second study, natural sensory signals were recorded from cutaneous nerves in the forelimb of six cats during walking on a motorized treadmill, using chronically implanted tripolar nerve cuffs on the Median, Ulnar, and/or Radial nerves.Features in the electroneurograms (ENGs) related to foot contact and lift-off were reliably detected and were used by a state controller model to predict the timing of activity of Palmaris Longus, a wrist flexor, for a variety of walking conditions.The cutaneous ENG signals recorded during walking suggested that natural sensory signals may be more generally implemented as a source of feedback for closed-loop control of FES.In a third study, a real-time state controller for FES was developed which implemented cutaneous neural signals from the cat forelimb as feedback to predict the timing of the stance phase and associated Palmaris Longus (PalL) muscle activity during walking on the treadmill.The state controller was tested in three sets of experiments: 1) to predict the timing of muscle activity in an open-loop configuration with no stimulation, 2) to predict the timing of muscle activity in a closedloop configuration and stimulate the PalL over a background of natural muscle activity, and 3) to predict the timing of muscle activity in a closed-loop configuration and stimulate the PalL to return function to the wrist during temporary paralysis of the PalL and other wrist planter flexors.The FES state controller was tested in a variety of walking conditions, including treadmill velocities of up to 1.0 rnls and slopes o f f 10%.The stability of nerve cuff signals over time and the accuracy of state controller results in a range of walking conditions lend support to the long-term use of nerve cuff recording electrodes for the control of functional electrical stimulation systems and neural prostheses in human applications.
Source: Summit (Simon Fraser University)
Author: K.D. Strange
Language: English
Subfield: Cellular and Molecular Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W2135961002