**4. Discussion**

Approximately 5.6% of all cochlear implant users report experiencing aberrant facial nerve stimulation (FNS) as a side effect of their CI implantation [4]. For users presenting with FNS, audiologists may first attempt to control the problem by re-programming the device to produce lower currents, followed by turning off offending electrodes—both of which can reduce speech comprehension [4]. If these solutions fail, clinics have observed that re-implantation with an Oticon Medical Neuro Zti implant can resolve FNS issues [8,10–13]. Indeed, for our two subjects, re-implantation with the Oticon Medical device not only completely resolved FNS but also improved speech recognition [8].

The mechanisms that underlie this improvement are not yet well understood. However, we can reasonably expect that factors that affect the local electrical fields near neural activation points for the auditory and facial nerves—combined with how those nerves respond to these fields—are involved. These factors include: (1) electrode proximity; (2) factors that affect current spread (e.g., grounding and pulse duration and shape); and (3) polarity. In the subsequent four sections, we explore these factors and discuss the distinct characteristics of the Oticon Medical device in these areas. It is important to clarify that our intention is to shed light on these differences and not to imply that these distinctions inherently make the Oticon Medical device superior to others.

#### *4.1. Electrode Proximity*

The proximity between the stimulating electrodes and the neural activation sites of the auditory nerve fibers (ANF) is influenced by the type of electrode array used. Common understanding suggests that modiolus-hugging or mid-scala arrays might offer advantages in minimizing FNS over the lateral wall array design, used—among others—in the Oticon Medical electrode array. Case studies, such as Battmer et al. [14], indicate that electrodes positioned closer to the ANF require less current for excitation. This reduced current potentially leads to limited current spread, decreasing the likelihood of stimulating more distant non-auditory neural structures. Indeed, when looking across the literature, electrode array type does emerge as a statistically significant factor [4]. However, other case studies—including those of the two subjects in this manuscript—demonstrate that the Oticon Medical device is effective in alleviating unwanted FNS. Consequently, as we've previously argued [8], stimulation-related factors likely have a larger impact.

#### *4.2. Grounding*

Beyond geometry, other factors determining the current spread are also relevant when considering the activation of more distant, non-auditory neural structures. A notable distinction between the OM devices and others lies in their DAP grounding scheme (see Figure 1). With DAP, approximately 80% of the current returns to intra-cochlear electrodes and the remaining 20% to an extra-cochlear electrode [7]. By contrast, conventional MPgrounding returns all the current through the extra-cochlear electrode. This MP grounding mechanism theoretically results in a broader dispersion of the overall electrical field, making it more likely to intersect with the facial nerve.

In addition to DAP and MP, 'bipolar' and 'common ground' schemes are also in clinical use, with the latter commonly observed in older Cochlear® (Cochlear Limited, Sydney, Australia) devices. Both return current via intracochlear electrodes. A study investigating the effects of different grounding strategies on FNS efficacy, conducted using 204 electrically evoked compound action potential (eCAP) input/output functions recorded from 33 ears of 26 guinea pigs, revealed that—for biphasic pulses—the broad-MP grounding was associated with a high occurrence of FNS (65%), while bipolar and an experimental tripolar configuration (expected to be the most focused) generated only 20% and 2% of FNS occurrences, respectively [15].
