East Carolina University
Place specificity of electrical stimulation with a cochlear implant and its relationship to neural survival and speech recognition

Modern cochlear implants code a speech signal by dividing it into spectral channels and modulating trains of biphasic electrical pulses with the low-frequency temporal envelope from each channel. Perception with a cochlear implant therefore is dependent on two factors: the acuity of processing the low frequency amplitude modulations and the place specificity of excitation. Cochlear implant users have demonstrated amplitude modulation detection similar to that found in normal hearing ears with larger individual variability, nonetheless the challenge of the prosthesis lies with the poor place specificity. Ideally, the output of an electrode should target a specific population of nerve fibers. However spread of excitation is often large, leading to neural interactions between channels. The first aim of the research is to investigate whether the underlying cause of channel interaction is the status of neural survival near the stimulation sites. The hypothesis is that excitation is more likely to spread further from a stimulation site in the cochlea when the density of surviving spiral ganglion cells is sparse. Neural status of a stimulation site will be estimated using a non-invasive psychophysical measure that correlates with the count of spiral ganglion cells in implanted animals. Place specificity of excitation will be measured by a psychophysical forward-masking procedure that is widely used to assess spatial tuning in electrical hearing. The second aim of the research is to examine the effects of place specificity on speech perception that requires mainly the spectral cues. The rationale is that poor place specificity would cause a reduction in spectral resolution or smearing of spectral speech cues. Proposed speech recognition tasks that demand good spatial tuning are perception of sine-wave speech, also known as the spectral skeleton of speech, and perception of speech presented in background noises.

Research area: Cochlear Implants

Long-term goal of research: If the proposed hypothesis is correct, that is, poor neural survival predicts poor spatial tuning, we will seek to ask the question whether the temporal processing acuity of the activated neurons at the site with low cell density would also be compromised. Once the relationship between temporal and spatial acuity is determined, we will examine the relative importance of the temporal and spatial acuity for speech recognition. The two acuities might not be equally important, or that they might be contributing to different aspects of speech recognition. With this information, the long term goal is to optimize implant user’s speech processing MAPs by strengthening the perception acuity that is the most important for speech recognition in the ear.