Acoustically derived auditory nerve action potential evoked by electrical stimulation: an estimation of the waveform of single unit contribution.

An experimental study of the electrical stimulation of the guinea pig cochlea is made using an electrode on the round window for both stimulation and recording. The neural response is separated from the electrical artifact with a masking procedure combined with a low amplification, "statistical" averaging method [Charlet de Sauvage et al., Hear. Res. 2, 343-346 (1980)]. The high electrical impedance required for recording physiological responses implies the use of a current pulse generator. Monitoring of evoked potentials from the auditory cortex provides evidence that the effects of electrical stimulation (and of masking noise) are of auditory origin. The electrically evoked round window response is of very short latency (less than 0.2 ms). There is a response threshold for both electrical stimulus and masking noise. The response amplitude varies monotonically as a function of masking noise or electrical stimulus intensity. Experiments with high-pass noise masking suggest that the electrical stimulus is mainly acting on basal fibers. The response latency and waveform are independent of electrical stimulus intensity, repetition rate, masker level, or spectrum. Little intersubject variation is noted. Our experiments (reciprocal forward masking by electrical and acoustic stimuli) suggest that a direct, instantaneous electrical stimulation of the fibers occurs. We believe that this response to electrical stimulation represents the mean unit response of the auditory nerve fibers. This approach may be useful in the separate study of cochlear and VIIIth nerve functions and in the analysis (deconvolution) of the acoustically evoked compound AP.