The multi-channel cochlear implant and the relief of severe-to-profound deafness

This personal reflection outlines the discoveries at the University of Melbourne leading to the multi-channel cochlear implant, and its development industrially by Cochlear Limited. My earlier experimental electrophysiological research demonstrated temporal coding occurred for only low frequencies, i.e. below 200-500 pulses/second. I was able to confirm these findings perceptually in behaviourally conditioned animals. In addition, these studies showed that temporal discrimination occurred across spatial coding channels. These experimental results correlated with the later conscious experience for electrical stimulation in my implant patients. In addition, the mid-to-high frequencies were coded in part by place of stimulation using bipolar and monopolar stimulation to restrict current spread. Furthermore, place of stimulation had the qualities of sharpness and dullness, and was also experienced as vowels. Owing to the limitation in coding speech with a physiological model due to the overlap of electrical current leading to unpredictable variations in loudness, a speech coding strategy that extracted the most important speech features for transmission through an electro-neural 'bottle-neck' to the brain was explored. Our inaugural strategy, discovered in 1978, extracted the second formant for place of stimulation, voicing for rate of stimulation, and sound pressure for current level. This was the first coding strategy to provide open-set speech understanding, as shown by standard audiological tests, and it became the first clinically successful interface between the world and human consciousness. This strategy was improved with place coding for the third formant or high-frequency spectrum, and then the spectral maxima. In 1989, I operated on our first patient to receive a bilateral implant, and in 1990, the first with a bimodal processor. The psychophysics and speech perception for these showed that the stimuli from each side could be fused into a single image, and localized according to differences in intensity and time of arrival of the stimuli. There were significant improvements for speech perception in noise. In 1985, I implanted our first children with the multi-channel prosthesis and found that speech understanding and spoken language were greatly improved the younger the child at surgery, and especially when younger than 12 months. Speech understanding was strongly related to the development of place coding. In 1990, the US Food and Drug Administration approved the implant for deaf children, the first by any world health regulatory body making it the first major advance in helping deaf children to communicate.