Cochlear implants (CIs) are neural prostheses that improve access to sounds for patients with severe to profound sensorineural hearing loss. CIs stimulate the auditory nerve with current pulses on a multi-electrode array that is placed inside the cochlea. Success of CIs varies based on a wide range of patient-related and implant-related factors. In this thesis, I investigate the use of CI electrical stimulation responses to assess CI status and the environment it occupies with regards to three issues that may affect CI performance and usage: extracochlear electrodes, facial-nerve stimulation, and cochlear fibrosis formation.

In chapter 2, I investigate the use of Stimulation-Current-Induced Non-Stimulating Electrode Voltage recordings (SCINSEVs), a non-invasive and quick measurement that does not require additional equipment, as a detection and quantification tool for extracochlear electrodes. The presence of extracochlear electrodes may cause poor CI performance through several mechanisms and commonly goes unnoticed. I show how SCINSEVs can be used, both in a human cadaver model and in CI patients, as a visual tool to detect extracochlear electrodes and I introduce an automated detection and quantification tool for those clinicians less familiar with SCINSEVs. This measure might be able to prevent revision surgery when used intra-operatively.

Chapter 3 discusses CI-induced facial-nerve stimulation, a form of non-auditory stimulation from which a subset of CI patients suffers, and which can severely limit CI use and performance. To understand CI-induced facial-nerve stimulation, I investigate the distribution of voltage in the facial nerve canal with intracochlear CI stimulation in fresh-frozen human cadaveric heads. I use the voltage distribution measured to model facial-nerve thresholds in a multi-compartment single fibre model of facial-nerve activation. I compare several pulse paradigms on their ability to stimulate the auditory nerve, modelled with a parametric finite-element model and an auditory-nerve activation model, without stimulating the facial nerve. The largest amount of ‘wiggle room’ between auditory-nerve mean comfortable loudness levels and facial-nerve thresholds was found for triphasic pulses.

In chapter 4, I present a tissue-engineered 3D in vitro model of fibrosis on a CI in which I culture fibroblasts in a fibrin gel wrapped around a CI electrode array. CIs are known to cause fibrosis formation intracochlearly, which can lead to residual hearing loss. I show that complex impedance can be used to track fibrosis formation over time by measuring electrochemical impedance spectroscopy (EIS) and voltage waveforms. Within the proposed circuit model in this chapter, the changes over time in this model can be attributed to a significant increase in the resistance and a decrease in the capacitance of the bulk of the gel, while no significant changes are seen at the level of the electrode-electrolyte interface. Based on this notion, I introduce the second phase peak ratio (SPPR) of voltage waveforms, which represents the ratio of resistance and capacitance, and show that they can be used as a marker of fibrosis development over time.

Overall, the work in this thesis adds to the clinical utility of CI electrical stimulation as a marker for extracochlear electrodes, facial-nerve stimulation, and cochlear fibrosis.