Using the voluntary EMG as a control signal for the stimulation of the same muscle necessitates elimination of stimulus artifacts and the muscle response caused by the stimulation. The stimulus artifacts are easily eliminated by shutting down the amplifier during stimulation. The muscle response is a non-stationary signal, therefore an adaptive linear prediction filter is proposed. The filter is derived and tested for three filter lengths on both simulated and real data. The performance is compared with a conventional fixed comb filter. The simulations indicate that the adaptive filter is relatively insensitive to amplitude variations og the muscle responses. For variations in shape and for real data, an improved filter performance can be achieved by increasing the filter length. A filter length of up to 7 stimulation periods makes it possible to reduce real muscle responses to a level comparable with the background noise. It is thus possible to extract the voluntary EMG from a partly paralysed muscle and use it for controlling the stimulation of the same muscle.
|Journal||IEEE transactions on rehabilitation engineering|
|Publication status||Published - 1997|