Controlled and selective activation of peripheral nerves using a multi -contact self -sizing spiral cuff electrode

Four contact, monopolar nerve cuff electrodes were successfully used to selectively activate each of four motor fascicles within a multi-fascicular nerve trunk and to produce the linear addition of the individual torque outputs. Present-day motor prostheses, using either electrodes on the skin surface or implanted on or in a muscle, require a separate electrode and lead for each muscle that is to be activated. As the motor prosthesis field advances, improved performance will require the selective control of many muscles. This work supports the hypothesis that a multiple monopolar contacts in a nerve cuff can activate selectively many muscles. Self-sizing spiral nerve cuff electrodes were used to deliver 10musec rectangular pulses to the cat sciatic nerve. The isometric torque about the ankle was measured in three dimensions: plantar flexion, lateral rotation, and inversion. In the first series of experiments six different animals were each implanted with a twelve contact nerve cuff electrode through which both tripolar and monopolar stimulation was applied. Both the torque output and the repeatability of monopolar and tripolar stimulation were found to be statistically within one standard deviation of each other. In the second series of experiments the torque outputs produced using a nerve cuff electrode with four monopolar contacts were compared to the torque outputs produced by individual cuff electrodes placed on each branch of the sciatic nerve. In each of the 10 cases attempted, multiple contacts were successfully used to achieve the force output of a fascicle not achieved with a single contact. Further, the torque outputs were shown to be a result of the full and selective activation of the corresponding fascicle. In the third series of experiments individual torque outputs produced by activation of different sets of nerve fibers were shown to add linearly when an appropriate delay between stimuli was applied. For each amplitude and location around each nerve tested, the range of delays between 700 and 900musec were shown to be appropriate. The results of this research support the hypothesis that a multiple contact, monopolar nerve cuff can be used to control the net torque output about a joint.