Cochlear implants (CIs) are neuro-prosthetic devices that bypass peripheral parts of the auditory system and give people with severe-to-profound hearing loss the ability perceive sound. However, many cochlear implant users struggle to hear in challenging listening environments, and there is a lot of variability between patients in their ability to understand speech. This highlights a need for better tools that provide patient-specific information about cochlear implant users, designed with the ultimate goal of enabling each individual implant patient to get the most out of their device and be able to understand speech as well as possible.

There are various behavioral / psychophysical tests that can provide insight into peripheral auditory neural health as well as methods for indirectly estimating current spread in the cochlea. However, many of these tests are time consuming and are not possible to run as part of routine clinical appointments where time is limited. The body of this thesis describes a different approach to determining patient-specific patterns of neural health and current spread along the cochlea using objective measurements of neural responsiveness.

The Electrically-Evoked Compound Action Potential (ECAP) is an objective measure of peripheral auditory neural responsiveness that can be measured in cochlear implant users. They are quick to record and are easy to do in the clinic as they require no additional hardware. They simply use the electrodes already present in the cochlear implant to stimulate and other electrodes also within the implant to record the neural activity. These have been used to estimate various attributes of the electrode-neuron interface, but have had mixed results when correlating with other metrics.

This thesis describes the development of an algorithm, the Panoramic ECAP Method, that uses ECAPs from individual CI patients to estimate patient-specific patterns of neural health and current spread. It discusses the theory and the maths behind the algorithm itself, various experiments done to validate the accuracy of the estimates, and methods developed to optimize the quality and the speed of the data collection to improve clinical viability.

The over-arching goal is to design an objective diagnostic tool that can be used in the clinic to provide patient-specific estimates of the electrode-neuron interface of cochlear implant users. This information may provide clinicians with information that can be leveraged to personalize cochlear implant programming to specific patients and help deliver auditory information as optimally as possible to their cochleae. These interventions may further be used to improve the speech perception for cochlear implant users who may otherwise struggle.