**1. Introduction**

Cochlear implant (CI) is the most effective treatment option for young children with profound sensorineural hearing loss. A longitudinal study showed that children implanted up to 2 years old scored on average above 50% on open-set speech recognition tasks after 4 years of CI experience [1]. The rate of complications with cochlear implantation has also decreased with advances in the CI field in the last years [2]. However, one potential complication that persists and affects the auditory performance of children with CI is the facial nerve stimulation (FNS).

**Citation:** Danieli, F.; Hyppolito, M.A.; Hussain, R.; Hoen, M.; Karoui, C.; Reis, A.C.M.B. The Effects of Multi-Mode Monophasic Stimulation with Capacitive Discharge on the Facial Nerve Stimulation Reduction in Young Children with Cochlear Implants: Intraoperative Recordings. *J. Clin. Med.* **2023**, *12*, 534. https:// doi.org/10.3390/jcm12020534

Academic Editor: Giuseppe Magliulo

Received: 7 December 2022 Revised: 28 December 2022 Accepted: 5 January 2023 Published: 9 January 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

FNS incidence was reported to range from 1.14% to 43% in children, with immediate or delayed onset [3]. Although it is known that otosclerosis, meningitis, temporal bone fractures and congenital cochlear anomalies increase the risk of FNS, some patients experience it after cochlear implantation without any of these etiologies. FNS symptoms may range from mild facial movements to severe facial spasms, painful or debilitating [4], either visually detected or self-reported by the patient. In young children, FNS has been underestimated, as they may not accurately report its symptoms. Cushing, Papsin and Gordon [5] reported a large difference in the incidence of FNS in children with CI when electrophysiological recordings are compared to their reports or even to observation of facial movements.

It is assumed that the electric current passing from the electrode to spiral ganglion cell can spread to the nearby facial nerve causing FNS [6], but the exact mechanism underlying the FNS remains unclear, as well as the relative contribution of factors to trigger the symptoms and the best treatment option to resolve it.

Some strategies have been adopted to manage FNS symptoms, including maximum comfort levels (MCL) reduction [7], pulse wide widening [8], the use of triphasic pulses [9], electrode deactivation [10] and cochlear re-implantation [11]. However, these strategies may result in auditory performance decline [7] and does not ensure to resolve FNS [8].

Recently, the use of the multi-mode monophasic stimulation was proposed as a promising strategy to manage FNS [7,12,13]. It resolved severe FNS in some adult CI recipients, after cochlear re-implantation with the Neuro Zti device (Oticon Medical, Smørum, Denmark). Most current CI devices use monopolar biphasic stimulation, and, in this CI electrical stimulation pattern, the total amount of electrical charge flows from intra-cochlear electrodes to extra-cochlear ground electrodes, and each phase of the pulse stimulates different group of neurons, increasing the spatial extent of stimulation. Using multi-mode monophasic stimulation with subsequent capacitive discharge, most of the electrical current is maintained within the cochlea, and the anodic stimulating phase of the monophasic pulse is followed by a non-stimulating cathodic phase (with reduced amplitude, compared to anodic phase). It is hypothesized that multi-mode monophasic stimulation decreases the spatial extent of electrical stimulation and reduces the amount of the current spread to the periphery structures, including the facial nerve, thereby reducing FNS.

To the best of our knowledge, the effects of multi-mode monophasic stimulation on FNS reduction in children were not previously investigated. Thus, in this study, we recorded intraoperative EMG action potentials to investigate the use of the multi-mode monophasic stimulation in young children and the effects of this stimulation pattern on the FNS reduction in this population. We also used 3D image processing techniques to estimate the CI intra-cochlear electrodes placement, as well as the distances between the basal turn of the cochlea and electrodes (based on their real intra-cochlear positioning) to the labyrinthine segment of the facial nerve, to investigate their influence on the EMG recordings.

#### **2. Materials and Methods**

This was an exploratory prospective observational study approved by the Institutional Ethics Committee under protocol 5.117.640. Parental informed consent was obtained from all subjects involved in the study.

#### *2.1. Subjects*

Ten ears from seven prelingually deafened children aged up to 6 years old who undergone either unilateral or bilateral CI surgeries were included in this study. All subjects were implanted with the Neuro Zti Evo® device associated to the Neuro 2 sound processor (Oticon Medical, Smørum, Denmark). The exclusion criteria was comprised of subjects with preoperative facial palsy or other facial nerve dysfunctions, neuromuscular diseases, and epilepsy, as they could affect the EMG responses. T demographic data of the subjects are shown in the Table 1.


**Table 1.** Subject demographics.

S1–S7: subjects 1–7; EA1–EA10: implanted ears 1–10; L: left; R: right; M: male; F: female; CI: cochlear implant; m: months.

#### *2.2. Procedure*

EMG responses were recorded in all subjects during CI surgery, under general anesthesia (Propofol and Fentanyl). The duration of measurement was about 10–15 min to not prolong the surgery time, and it was carried out immediately after the insertion of the EVO® electrode array inside the cochlea of the subjects. Pre-anesthetic sedatives were not administrated to avoid muscle relaxation and their influence on the EMG recordings and facial nerve monitoring.

FNS stimulation was investigated through the EMG action potentials recorded from the orbicular oculi or oris muscles, both innervated by the facial nerve. Prior to sterile surgical preparation, bipolar needle electrodes were placed inside the orbicularis oculi and oris muscles, ipsilateral to the cochlear implantation site. Intraoperative facial nerve monitoring was initially performed to assess EMG recordings from inputs (i.e., orbicularis oculi and oris muscles) using the Nerve Integrity Monitor—NIM-2 equipment (Medtronic Xomed Inc., Jacksonville, FL, USA). After insertion of the electrode array, the experimental protocol was performed firstly using only the orbicularis oculi input channel, to limit the duration of the measurement in young children. If absent or no clear responses were recorded from this channel, the orbicularis oris input channel was then used.

#### *2.3. Stimulation Parameters and EMG Responses*

The intracochlear electrical stimulation was produced by the cochlear implant, using the eCAP tool of the Genie Medical CI fitting software, version 1.6, and CI-Link interface (Oticon Medical, Smørum, Denmark), connected to a computer and external antenna, responsible for transmitting the electrical stimuli to internal antenna via radiofrequency. Four electrodes were tested in each ear: one basal (E1), two medial (E8, E15) and one apical (E20). For the facial nerve stimulation thresholds (T-FNS) investigation purposes, current levels started from 20 SA (stimulation amplitude, 1 SA = 1/45 mA) and increased by 5 SA steps until reach 70 SA (maximum current level). The pulse duration was fixed at 30 SD (stimulation duration, 1 SD = 1 μs). The stimulation level (nC/phase) at which an EMG response was first evoked was defined as T-FNS, and no further increase in stimulus level was performed after detection of an FNS response. Peak-to-peak amplitudes of EMG responses at T-FNS were also recorded.

In order to investigate EMG responses using the two different CI stimulation patterns, the experimental protocol first employed monopolar biphasic stimulation (ST1) and then, multi-mode monophasic stimulation (ST2). Figure 1 shows the schematic of different stimulation patterns used in this study. Stimulation parameters used to record EMG responses are provided in the Table 2.

**Figure 1.** Schematic representation of different stimulation patterns used in this study. (**A**) Monopolar biphasic stimulation (stimulation pattern ST1) and (**B**) multi-mode monophasic stimulation with capacitive discharge (stimulation pattern ST2).



CD: capacitive discharge; MP: monopolar; MM: multi-mode grounding, min: minimum; max: maximum.

#### *2.4. Radiological Examination*

Pre- and postoperative CT scans of the ears were performed to investigate the intracochlear electrodes placement and the distance between the cochlea and labyrinthine segment of the facial nerve (cochlea-nerve distance). For this purpose, postoperative CT scans were performed in all subjects three months after cochlear implantation.

The CT-scans were acquired at a resolution of 0.3 × 0.3 × 0.4 mm3. CT image reconstruction was performed using a web-based research platform Nautilus [14], which combines a convolutional neural network (CNN) approach with Bayesian joint appearance and shape inference for segmenting the cochlea and determining the trajectory of the electrode arrays. The angular positioning of the electrodes was determined automatically using Nautilus which extracts electrode position from postoperative CT-scan using a CNN approach and registers it with preoperative CT-scan for determining the electrode position with respect to the cochlear segmentation. The closest distance between the lateral wall of the cochlea and the labyrinthine segment of the facial nerve (cochlea-nerve distance) were measured on the pre-operative CT-scans. Next, the cochlea-nerve distance at each electrode's angular location (electrode-nerve distance) were also measured to estimate the electrodes placed closest to the facial nerve. The 3D cochlear view reconstruction, including scala tympani segmentation, was input to Slicer 3D and manual annotation of the facial nerve was performed (Figure 2).

**Figure 2.** Model of CT image reconstruction performed by the software Nautilus in this study: cochlear segmentation, angular positioning of the electrodes and cochlear-nerve and electrode-nerve distances estimation.
