Signal processing concepts for optimal myoelectric sensor placement in a modular hybrid FES orthosis

Abstract

The project "OrthoJacket" (BMBF grant 01EZ 0774) aims for the development of a novel hybrid orthosis based on external fluidic actuators and Functional Electrical Stimulation (FES) for functional restoration of paralyzed upper extremities in case of high cervical spinal cord injuries (SCI). The main challenges of the realization of this system are the mechanical stabilization of the segments of the upper extremity, the generation of precise and graduated joint movements and the detection of the users' movement intention. Since in many tetraplegic persons muscle functions are preserved to a certain extend, electromyographic (EMG) signals seem to be best suited as a control source of the hybridorthosis. However, in most cases the preserved muscle activities are very weak and therefore the determination of the best position for placement of the EMG sensor is most crucial for achieving a sufficient signal-to-noise ratio. In this paper we present a method for determination of the optimal EMG sensor position in terms of signal amplitude and selectivity. The proposed method is based on a quality criterion, which is derived from the EMG response of the muscle to electrical stimulation pulses. In a pilot study with one healthy subject and three predefined sensor sites we have been able to show the basic functionality of the method. © 2009 Springer Berlin Heidelberg.