Automatic and individually optimised fitting of cochlear implants

A cochlear implant is a device to help deaf people hear. It translates sounds into electrical impulses applied to the hearing nerve in the ear. It is an important device that allows spoken communication and, for infants, the opportunity to develop spoken language. This project has two broad aims. The first is to develop ways to fit a cochlear implant without having to do lengthy tests involving the patients listening and making judgments about sounds. We will do this by directly measuring the activity of the hearing nerve and the brain in response to sounds. The second broad aim is to understand why different people are more or less successful with the implant, and, armed with this information, how to individually optimise the fitting so that the potential of each person is fully realised. Once again we will measure the responses of the hearing nerve and the brain to sounds, and use these to understand the differences in the way that each person hears with a cochlear implant. We can then find the significance of these differences for speech understanding and work out ways to optimise the settings of the sound processor based on individual characteristics. In summary, it is hoped that the research project will improve speech understanding with a cochlear implant, especially for those who currently do not do as well as their peers. Importantly, it is hoped that good quality fittings can be achieved for infants thus maximising their potential for spoken language development.

This project has two broad aims: to improve methods of automatic fitting of cochlear implants using objective measures; and to use objective measures to assess individual differences in neural function that may be relevant to selection of optimal sound processing strategies.
We will use a combination of techniques which include psychophysical tests, mathematical modelling of hearing, speech perception tests, and electrophysiological measures of auditory pathways. Hypotheses are tested in cohorts of 20 adult cochlear implantees, and, where appropriate, methods will be further evaluated in a cohort of infants who use a cochlear implant. The objective measures used will be electrically-evoked compound action potential (ECAP), assessing function of the peripheral auditory nerve, long-latency auditory evoked potential (LAEP), including the acoustic change response (ACR) that reflects cortical response to a change in stimulus, and the electrically-evoked stapedial reflex (SR), a middle-ear muscle reflex to loud sounds that involves the brain stem auditory pathways.
The first three experiments aim to develop improved objective measures to determine the range of electrical currents used on each electrode without the use of behavioural data. The first two experiments explore ways of using ECAP and LAEP potentials respectively to determine the electrical current threshold levels on each electrode needed for fitting. In both cases the mathematical loudness model for electrical stimulation of McKay et al. is used in the development of the hypotheses and in the data analysis. The third experiment explores the relationship between the SR threshold and loudness using the same loudness model, and assesses the use of SR to determine the upper limit of electrical current used on each electrode for fitting the implant.
The last two experiments investigate the individual differences in peripheral and central neural response behaviour respectively, and seek relationships between these differences and perceptual ability. In the fourth experiment we test a specific hypothesis that the rate of stimulation which optimises speech understanding is related to the degree of peripheral neural survival or the functional refractory characteristics of the surviving neurons. In the final experiment we explore electrophysiological analogues of perceptual temporal and spectral resolution and determine their importance for speech understanding.
The overall outcome is hoped to be an improved and individually-optimised method of fitting patients using only objective rather than behavioural data. Improved fitting (and earlier fitting in infants) will lead to better outcomes, and hence quality of life, for people with a severe or profound hearing loss.