Cochlear implants (CIs) can restore hearing not only in patients with severe to profound hearing loss and deafness, but also in patients following tumour removal of intracochlear schwannomas, that are benign. These tumours originate in the peripheral branches of the eighth cranial nerve and often require surgical excision, resulting in partial or subtotal cochlear removal [1]. Interestingly, affected patients who were treated with CIs can potentially demonstrate comparable outcome performance as single-sided deafened CI users [1,2].
To evoke a sound perception in CI users, cochlear nerve fibres are stimulated by electrical current pulses along the electrode array. The amount and location of the excited nerve fibres depend on several parameters, such as the relative position of the electrode array in the cochlea, electrical impedances of the electrode-tissue-nerve connections, respective nerve sensitivity, parameters of the stimulation pulses, and more. Without doubt, these factors influence intracochlear current flow and impede precise steering altogether, thus the resulting spread of excitation of nerve fibres can span about one third of the electrode array or even more [3,4]. Unfortunately, this undifferentiated distribution eventually leads to excitation of unaimed nerve fibers and thus to inaccurate auditory representation - still a limiting factor in current CI technology. Therefore, the enhancement of speech and music perception based on improved spatial resolution in CIs is widely discussed [5,6].
The CI induced current distribution and flow within the cochlea was modelled in several studies (see review 7), which is based on different influencing factors, e.g., electrical resistivity of the bone and tissue [8,9], electrical conductivity of the modiolus encapsulation [8], and changes in electrode-electrolyte interface by charge injection [7,10]. Instead, Choi et al. [11] used finite elements to model electrical impedances across the electrode array based on the assumption that the distance between electrode array and modiolus stays constant along its progression. However, the validity of this assumption has been questioned in the following studies [12,13]. However, using these insights, Mens [14] was able to show strong gradient related field distributions, where the apical to basal voltage propagation shows an increase in amplitude of about three times, thereby strongly emphasizing the necessity of further investigations into such mechanisms and effects.
In addition to electrochemical and physiological factors one can hypothesize that the actual design of the CI electrode array influences the spread of electric field. During the last decades CI manufacturers developed and offered a variety of different electrode designs, e.g., long lateral wall arrays with the aim to cover the complete cochlear length and nerve fibres (MED-EL, Austria), shorter perimodiolar electrode arrays to decrease distance between carrier and nerve fibres (Cochlear, Australia), and others. Study results comparing the difference in word recognition between perimodiolar and lateral wall electrode designs finally base on the different electrical conditions within the cochlea show facilitated word recognition in perimodiolar arrays [6].
However, all these findings are based on the fact that the underlying anatomy of the cochlea is intact and its compartments remain fluid-filled. In contrast, in tumour patients with partial or subtotal cochleoectomy, the fluid-filled compartments are degraded and thus their contribution to the electric field distribution is impaired. Consequently, these patients could particularly benefit from precise control of the electric fields and thus from differential recruitment of the nerve fibres. For those patients, a perimodiolar electrode array was developed by Med-EL (Innsbruck, Austria) as a custom made device (CMD) based on the template of their common FORM19 electrode array [15].
To measure electrical impedances and voltage distributions CI manufacturers offer system integrated telemetry functions, that are commonly used in daily clinical routine to monitor correct implant functioning. One particular telemetry feature is the measurement of the intra-cochlear electric field or intra-cochlear current spread (ICCS) [16], labelled different for different manufacturers electrical field imaging (EFI) or transimpedance matrices (TIM) [17] or voltage matrices. All these tools measure the impedance and voltage gradient along the electrode array for the respective CI system, where values can then be used to calculate the voltage gradient along the nerve fibres. In MED-EL devices this measure is included in each routine impedance measure.
Interestingly, significant peculiarities in the electrical field distribution were shown across different patient groups. For instance, Wagner et al. [2] observed smaller electric fields in Nucleus CI patients after partial or subtotal cochleoectomy for removal of intracochlear schwannoma compared to patients who underwent standard cochlear implantation through a round window approach. Noteworthy, patients who underwent cochleoectomy at cochlear implantation were able to demonstrate comparable or even better outcome in speech perception compared to the group with “normal” CI insertion [1]. Da Silva et al. [18] found that more precisely differentiated areas of recruited nerve fibres correlate with better outcomes in speech perception using spread of excitation measurements. However, the exact effects of stimulation field distributions are far from being understood.
Therefore, it is of deeper interest of this study to further investigate the relationship of individual electrical field distributions with speech perception results. To facilitate a robust fundament for analysis two different types of electrode arrays were used, one group of commonly used electrode arrays as well as one group of special CMD electrode arrays (both provided by MED-EL), to provide capability in comparison based on their diverging designs. Further comparability is provided by the analysis of the discussed special patient group who underwent cochleoectomy and were inserted using the CMDs. Fortunately, this combination offers an innovative and rare approach in research regarding this topic as these special arrays are one of a kind worldwide. Noteworthy, at the time of implantation only MED-EL implant technology allowed for postoperative MRI diagnostics, what was essential for the follow-up therapy of tumour patients. In summary, we expect to find differences in provided field distributions among the two investigated arrays based on their design and patient related anatomy as well as deeper insights into the relationship to speech performance outcomes.