Research Focus

The cochlear implant is probably the most successful bioengineering product ever. This is a device that electrically stimulates the auditory system of profoundly deaf individuals to elicit a sensation of sound. Tens of thousands of profoundly deaf people worldwide have regained some hearing by using cochlear implants. These are people that could not benefit from hearing aids, but have now regained the ability to use the telephone. 

The implanted part of a cochlear implant is an array of electrodes that is inserted in the cochlea during surgery. The outer part consists of a microphone and a speech processor that transforms and encodes the sound signal. The encoded sound is transmitted to the implanted electrodes by a radio signal. The electrodes then electrically stimulate nerve fibres in the cochlea to elicit a sensation of sound.

From the clinical viewpoint, cochlear implants are safe, reliable, and provide adequate sound information to deaf people. Despite being so successful, cochlear implant systems do not sound like normal hearing. Some users describe the sound as being similar to a radio that is tuned off the station. Not all users perform equally well and some users derive very little benefit due to poor quality of sound. Even the best implant listeners have difficulty in following speech in noisy conditions, and most implant users do not like the sound of music.  

Ideally, cochlear implants must emulate normal hearing. Most of the current research on cochlear implants worldwide revolves around the problem of how to achieve improved sound quality. Our research group does research that aims to improve understanding of the relationship between the design of the cochlear implant and the perceived sound.  

Despite the success of cochlear implants as seen from a clinical viewpoint, this relationship is not well understood. Research in our research group includes computer modelling of cochlear implants and psychoacoustic experimental work. One of the valuable contributions is a very flexible model of the implanted cochlea (see figure 1). This model is used to explore the influence of different designs and placement of cochlear implant electrodes in targeting the stimulation of specific regions of nerve fibres. This in turn aids understanding of how changes to different aspects of the cochlear implant design may result in improved sound quality.

Figure 1. This figure shows a computer model of a section of the human cochlea (a spiralling inner ear structure) in red, with nerve fibres in green. A section through two different placements of electrode arrays can be seen (in blue or red) at both ends of the spiral.

We have been developing three areas of work targeted at research questions in cochlear implants. The first involves the electrode: the design, the electrical fields around it, and the nerve fibres targeted by the electrodes. This is investigated using finite element models.

The second is to understand the relationship between stimulus and perception. This is investigated by creating computer models (that exist primarily in Matlab) of the signal processing in the normal auditory system and the electrically stimulated auditory system. The modelling work is supported by psychoacoustic experimental work involving individuals with normal hearing as well as cochlear implant users.

The third area of research aims to enable normal-hearing people to listen to the sounds that cochlear implantees hear. To this end, acoustic models of cochlear implants are created. Our hope is that appropriate acoustic models may assist researchers to develop improved speech processing algorithms.