Multiple-channel stimulation of the cochlear nucleus.

To further test the feasibility of a central nervous system auditory prosthesis, the characteristics of the electrically evoked middle latency response were studied in a series of acutely anesthetized pigmented guinea pigs, with multi-channel penetrating cochlear nucleus electrodes placed into the cochlear nucleus under direct visualization. These stimulating electrodes consisted of a silicone substrate, with five stimulating pads each, sputtered with iridium. Monopolar and bipolar stimulation were used. Threshold, latency, and input-output functions of the electrically evoked middle latency response were studied. Systematic differences were observed, depending on the site and parameters of stimulation. Principally, higher currents were required to produce waves of equal amplitude when the electrodes were closely spaced. For near electrode pairs, the maximum wave amplitudies obtainable within the limits of tissue safety were much lower than for distant electrode pairs. The slope of the growth function curve was steeper for widely spaced electrodes than for adjacent sites. Monopolar stimulation demonstrated maximum wave amplitudes with the lowest current intensity, implying current spread to the entire cochlear nucleus with this stimulation montage. In some cases, threshold differences were observed, higher thresholds being associated with closely spaced electrodes. These findings are consistent with simple models of the electric fields expected to be generated by these electrode arrays. The results support the hypothesis that activation of subpopulations of auditory brainstem neurons with multi-channel penetrating microelectrodes is possible.