Models of the peripheral nerves for detection and control of neural activity

Functional electrical stimulation (FES) can restore volitional motion of patients with neurological injuries or diseases using electrical stimulation of nerves innervating the muscles to be controlled independently. The Flat interface nerve electrode (FINE) enables the selective control of different muscles at the same time. In addition, multiple contact electrode designs allow selective recording of the various signals within the cuff. However, motion control of neuromuscular skeletal systems using multi-contact electrodes is a challenging problem due to the complexities of the systems and the large number of channels required to activate the various muscles involved in the motion. The localization and the recovery of many signals pose a significant challenge to the low signals to noise ratio and the large number of fascicles. Using computer models of the peripheral nerve, we have tested the ability of various algorithms to control the neuromuscular skeletal dynamics. Computer models have also been used to develop new methods to recover fascicular signals within the nerve. Both the control and the detection algorithms are currently being tested experimentally and preliminary results are included. The goal of this study is to develop the ability to detect nerve signals and use these signals to control joint motion in patients with stroke, amputation or paralysis.