Electrical stimulation of the auditory nerve: II. Effect of stimulus waveshape on single fibre response properties

Abstract

To investigate the generation of action potentials by electrical stimulation we studied the response of auditory nerve fibres (ANFs) to a variety of stimulus waveforms. Current pulses were presented to longitudinal bipolar scala tympani electrodes implanted in normal and deafened cochleae. Capacitively coupled monophasic current pulses evoked single ANF responses that were more sensitive to one phase (the ‘excitatory’ phase) than the other. Anodic pulses produced a significantly shorter mean latency compared with cathodic pulses, indicating that their site for spike initiation is located more centrally along the ANF. The fine temporal structure of ANF responses to biphasic pulses appeared similar to that evoked by monophasic pulses. An excitatory monophasic pulse evoked a significantly lower threshold than a biphasic current pulse having the same polarity and duration leading phase, i.e. the addition of a second phase leads to an increase in threshold. Increasing the temporal separation of the two phases of a biphasic pulse resulted in a moderate reduction in threshold which approached that of an excitatory monophasic pulse for interphase gaps >100 μs. Greater threshold reductions were observed with narrower current pulses. There was a systematic reduction in threshold with increasing pulse width for biphasic current pulses, reflecting the general charge-dependent properties of ANFs for narrow pulse widths. Chopped biphasic current pulses, which uniformly delivered multiple packets of charge (2×30 μs, 3×20 μs or 6×10 μs) with the same polarity over a 120 μs period, followed by a similar series in the reverse polarity, demonstrated the ability of the neural membrane to integrate sub-threshold packets of charge to achieve depolarisation. Moreover, thresholds for these current pulses were ∼1.5 dB lower than 60 μs/phase biphasic current pulses with no interphase gap. Finally, stimulation using charge-balanced triphasic and asymmetric current pulses produced systematic changes in threshold and latency consistent with the charge-dependent properties of ANFs. These findings provide insight into the mechanisms underlying the generation of action potentials using electrical stimuli. Moreover, a number of these novel stimuli may have potential clinical application.

Introduction

The previous papers in this series have described electrically elicited auditory nerve fibre (ANF) and auditory brainstem responses in normal, short-term deafened and long-term deafened cats (Shepherd and Javel, 1997), and have detailed ANF responses from these animals in terms of categorising the site of action potential generation with manipulations in stimulus intensity, waveform polarity and stimulus rate (Javel and Shepherd, 1999). In the present paper we describe the response of ANFs recorded from these animals to stimulation using a variety of stimulus waveforms including monophasic, biphasic, triphasic, chopped, and asymmetric current pulses, as well as temporal manipulations in both pulse width and interphase gap.

Although there have been a number of studies describing ANF responses to a variety of stimulus waveforms, the range studied has been relatively limited. Studies using monophasic or charge-balanced biphasic current pulses have shown that the ANF response is dominated by a highly synchronised short-latency (∼0.5 ms) response (Moxon, 1967, Moxon, 1971, Kiang and Moxon, 1972, Hartmann et al., 1984, van den Honert and Stypulkowski, 1984, van den Honert and Stypulkowski, 1987a, van den Honert and Stypulkowski, 1987b, Javel et al., 1987, Parkins and Colombo, 1987, Parkins, 1989, Hartmann and Klinke, 1990, Javel, 1990, Dynes, 1996, Shepherd and Javel, 1997). Studies using sinusoidal, triangular or square wave stimuli have indicated that threshold is frequency-dependent with the lowest thresholds occurring at ∼100 Hz and increasing with stimulus frequency, and that action potentials can be elicited by either the anodic or cathodic phase of the stimulus waveform (Hartmann et al., 1984, van den Honert and Stypulkowski, 1987a, Parkins, 1989).

An increased understanding of the response of ANFs to manipulations in stimulus waveshape has potentially important clinical implications. While biphasic current pulses are almost universally used clinically, from an electrochemical point of view other short duration charge-balanced current pulses would also appear to be suitable (Brummer and Turner, 1977, Robblee and Rose, 1990). However, little is known of the effects of these stimuli on ANF response characteristics. Potential clinical applications include: (i) the use of an interphase gap to increase sensitivity of biphasic current pulses (motor nerve: van den Honert and Mortimer, 1979; auditory nerve: P.M. Seligman, personal communication); (ii) employing chopped biphasic current pulses to effectively deliver charge to two or three cochlear sites at one time without the potential adverse effects of simultaneous current interaction associated with the vector summation of current density (McDermott, 1989); (iii) using triphasic current pulses to minimise interactions as a result of non-simultaneous stimulation by restoring the neural membrane potential to the resting state rather than leaving residual charge on the membrane (Eddington et al., 1995), and (iv) the use of asymmetric charge-balanced biphasic pulses to provide improved spatial selectivity (Frijns et al., 1996).