ABSTRACT

For severely paralyzed individuals communicating with their assistive technologies is often incompatible with their social interactions. To help remedy this, we investigated the extraction of two control channels from the surface electrical signals (sEMG) of a single muscle, which may be used in a variety of human-machine interfaces. We simulated sEMG signaling to help understand the best ways to extract multiple separate simultaneous control channels. We then acquired the sEMG signals of a facial muscle in 4 able-bodied subjects and trained them to place computer cursors at different (X,Y) spots on a computer screen by simultaneously generating different powers in two frequency bands. This method could be adapted to allow simultaneous control of multiple degrees of freedom of a single device, such as an environmental control device or a powered wheel chair.