Temporal representations with cochlear implants

Objective: To record and characterize intracochlear evoked potentials (EPs) for a variety of electrical stimuli in studies with cochlear implant patients. Methods: Recordings were made with patients having direct percutaneous access to their implanted electrodes. Intracochlear voltages were recorded via unstimulated electrodes. The stimuli included trains of identical pulses, with pulse rates ranging from 1()0 to 4065/s, and a modulated pulse train produced by a single-channel speech processor, with the pulse rate of 824/s. Results: Magnitudes of EPs for each pulse in trains of identical pulses were uniform for pulse rates below about 200/s. For rates between about 400 and 1000/s, an alternating pattern of EP magnitudes was observed, with relatively large EPs following the odd-numbered pulses. For rates between about 1000 and 3000/s, more complex patterns were observed. After the first millisecond of each train at even higher rates, uniform EPs again were observed across pulses, although the absolute magnitude of the EPs was much lower than that observed for low rates of stimulation. The approximate rates corresponding to boundaries between these different regions varied among subjects and among electrodes within subjects. EP magnitudes for the modulated pulse train reflected the gross periodicity of the modulation waveform but did not reflect temporal details within the periods. Conclusions: Population responses of the human auditory nerve, as indicated by EP magnitudes, reflect the amplitudes of electrical pulses for pulse rates below about 200/s and abouve about 3000/s. Use of intermediate rates may introduce distortions in the transmission of stimulus information with cochlear implants.