Quantification of muscle-derived signal interference during monopolar needle electromyography of a peripheral nerve interface in the rat hind limb

Abstract

High-fidelity signal acquisition is critical for the fundamental control of a neuroprosthesis. Our group has developed a bio-artificial interface consisting of a muscle graft neurotized by a severed nerve in a rat hind limb model. This regenerative peripheral nerve interface (RPNI) permits nerve signal transmission, amplification, and detection via in situ electromyography (EMG). Our study examined the magnitude of signal interference from simultaneously contracting muscles adjacent to our muscle of interest. In eighteen F344 rats, the extensor digitorum longus (EDL) muscle was used to fabricate simulated RPNI constructs of various sizes in which the neurovascular pedicle was preserved, obviating the need for reinnervation or revascularization. After 3 weeks of recovery, in situ EMG testing was performed using electrical stimulation of the common peroneal nerve. A recording needle was placed in the EDL muscle with a reference/ground electrode in the contralateral toe webspace, comprising a monopolar recording configuration. The superficial peroneal nerve was transected to further isolate stimulation of the anterior compartment. Recordings from the EDL were performed before and after excision of the tibialis anterior (TA) and extensor hallucis longus (EHL) muscles. After TA/EHL excision, EDL compound muscle action potential (CMAP) peak-to-peak amplitudes were significantly lower by an average of 7.4±5.6(SD) mV, or 32±18%, (paired t(17)=-5.7, p<0.0001). A significant positive linear correlation was seen between CMAP amplitude and EDL mass both before TA/EHL excision (r=0.68, n=18, p<0.01) and after TA/EHL excision (r=0.79, n=18, p<0.0001). EDL mass did not correlate with differences in CMAP amplitude or area caused by TA/EHL excision. Monopolar needle EMG recordings from the EDL muscle are significantly, but predictively, contaminated by concomitant muscular contractions in the anterior compartment of the rat hind limb. Further investigation of strategies to reduce this signal interference, including electrode choice or configuration, use of bioelectrical insulators, and filtering methods, is warranted to promote high-fidelity signal acquisition for prosthetic control.