Characterization of motor nerve stimulation using sinusoidal low frequency alternating currents and cuff electrodes

Objective:Direct electrical neurostimulation using continuous sinusoidal low frequency alternating currents (LFAC) is an emerging modality for neuromodulation. As opposed to the traditional rectangular pulse stimulation, there is limited background on the characteristics of peripheral nerves responses to sinusoidal LFAC stimulation; especially within the low frequency range (<50 Hz). In this study, we demonstrate LFAC activation as a means to activate motor nerves by direct bipolar nerve stimulation via cuff electrodes, and characterize the factors of activation. We study and quantify the effects of sinusoidal frequency and electrode geometry on the motor nerve activation thresholdin-vivoand in computational models,in-silico.Approach:Acutein-vivoexperiments (N= 34) were conducted on isoflurane-anaesthetized rats. A pure tone continuous sinusoidal current was applied to the rat sciatic nerve in bipolar configurations via bipolar or tripolar nerve cuff electrodes (different contact separations). LFAC activation thresholds were quantified by measuring the electromyogram (EMG) response of the triceps surae muscles and the induced twitch force to LFAC stimulation at six frequencies (1, 2, 3, 4, 8, and 20 Hz). Computationally, we utilized a volume conductor model of a bipolar cuff electrode around a single rat-size fascicle and projected the potentials to the McIntyre-Richardson-Grill models of myelinated motor nerve fibers. We compared thein-silicoresponses of a range of fiber diameters (5.7 to 16µm) to LFAC stimulation and their activation thresholds to thein-vivofindings.Main results: Sinusoidal LFAC stimulation elicited motor nerve activityin-vivoandin-silico, with a remarkable convergence of thein-silicopredictions to thein-vivoobservations. The EMG activity showed that muscle responses to LFAC stimulation were phase-locked to the sinusoidal cycle but exhibited two distinct activation modes. These modes were classified as burst and unitary, indicating the presence of two distinct patterns of muscle activation during LFAC stimulation. The LFAC motor activation threshold was significantly dependent on frequency and influenced by the contact separation of the cuff electrode, with a greater extent of reduction at a higher frequency or wider separation. Moreover, the order of fiber recruitment was suggested to be normal-physiological (small-to-large caliber) given the nature of the induced EMG activity andin-silicopredictions.Significance: These findings provide significant insights into the nature of sinusoidal LFAC stimulation, at the explored range of frequency, and the expected mammalian peripheral motor nerve responses to LFAC. The characteristics of sinusoidal LFAC stimulation would facilitate selectivity approaches in a broader range of therapeutic and rehabilitative neuromodulation applications.