The benefit provided by additional electrodes in medical implants is currently unknown. A typical cochlear implant only contains 12-22 electrodes to stimulate about 30,000 auditory nerve fibres - a task that is normally achieved by several thousand inner hair cells. This project will use computational modelling of anatomy and auditory physiology to develop a method to predict the performance of electrode numbers and stimulation schemes in terms of neurolinguistic function.
This project will build on existing FEA anatomical models of the cochlea and surrounding structures and will integrate these with differential models of neuron depolarisation such as the Hodgkin-Huxley model and the Meddis inner hair cell model. It will also develop literature-based models of speech and sound cognition which use the modelled neural response as their input. This "end to end" model will be allow performance to be predicted as a function of number of electrodes as well as other factors likely to influence cognition such as cochlear anatomy, state of neural survival, location of electrodes in the cochlea etc. It may be extended to embedded optimisation of computation in digital or analogue electronics. The model may be verified through measurements on cochlear implant recipients.
Web link for School of Electrical and Information Engineering: http://www.ee.usyd.edu.au
The opportunity ID for this research opportunity is 1757