Mathematical modelling of the electric potential from cochlear implants for a new diagnosis tool

Cochlear implants restore hearing in people with hearing loss, by replacing damaged or missing sensors in the cochlea in the inner ear. However, some people experience persistent problems with their implant. Precise diagnosis of the cause of these problems is often not possible, resulting in people with poor hearing despite having an implant.

Hearing loss is a significant risk factor for depression, dementia and disability-affected life-years with an estimated cost to the UK economy of >£40billion/annum. Cochlear implants could benefit many more deaf people and help to address a key societal challenge: a reduction in the burden of disability in later life. However, for this to happen, we need new diagnostic approaches to ensure lifelong good quality hearing for people with implants.

This project will transform the diagnosis of problems in cochlear implants by applying mathematics to develop a diagnostic tool that is driven by and designed for use in the clinic, that is easy to administer and that is well tolerated by patients.

A cochlear implant is a series of metal electrodes, which are inserted into the cochlea. Sound is captured by a unit on the ear that is connected to an internal stimulator on the skull. The stimulator causes the electrodes on the array in the cochlea to emit electric currents which activate the auditory nerve, sending signals to the brain which are recognised as sound.

Problems can occur that disrupt the electric current and the signal relayed to the brain, causing poorer hearing or loss of sound. The current from the electrodes also causes voltages that can be detected on the scalp. We have measured these voltages in people with cochlear implants and found that they can identify problematic cases where the implant is no longer relaying sound as expected.

Disruption to the electrical signals can also cause pain, or unpleasant sensations. These symptoms are distressing and hard to explain or treat, particularly when the underlying cause is not clear.

We aim to build a mathematical model of cochlear implants operating in the human head using finite element methods: these methods are widely applied in many areas of medicine and industry. The model will predict the voltages on the scalp that are generated in response to each implant electrode producing a current in turn. We will compare these predictions to data from people with cochlear implants to validate the model, using existing data as well as collecting new data.

To ensure the model can be applied to different people with differently shaped heads, we will study the effects of varying characteristics of the model head, such as its shape, capturing the right level of detail in our model.

To diagnose problems, we will study the relationship between each electrode and the voltage at specific locations on the scalp. We will describe this relationship for fully functional implants, and will then incorporate common problems, such as misplaced implants or build-up of scar tissue, into our model. Preliminary results show that the relationship between which electrode produces the current and the voltage on the scalp changes significantly in the presence of such problems.

This lays the foundation for using these relationships to diagnose problems. Our work focuses on the development of a clinical test that can be used on all patients, whereas prior work has generated detailed models of a few individuals.

Cochlear implants have a lifespan of 20-25-years, and problems can reduce this lifespan or result in additional surgery being needed. This is detrimental both to the patient and health budgets. There is a clear need for a reliable diagnostic tool that is quick to run in a standard clinical environment and usable in adults and children, irrespective of language skills or cognitive abilities.

This project paves the way for the development of a validated test that can be used in all cochlear implant clinics and ultimately, via telemedicine.