Evaluation of an impedance-based method to monitor the insertion of the electrode array during cochlear implantation

This study conformed to the Declaration of Helsinki and the guidelines of the Institutional Ethics and Research Committee of the Medical University of Innsbruck, which approved the study. Informed consent was obtained from all participants.

Between 2021 and 2022, a total of 20 patients (13 males, 7 females) who underwent cochlear implantation from 20 to 88 years of age (mean: 62 years, STD: 17 years) participated in the study. They were all implanted with an MEDEL Synchrony II® Cochlear Implant (CI) (MedEL, Innsbruck, Austria) with Flex electrode array (1 Flex 24, 18 Flex 28, 1 Flex Soft). Eleven patients received a CI on the right side and nine on the left side (Table 1). Each patient underwent a pre-operative and post-operative CT-Scan.

The Insertion Monitoring Tool (IMT) research software was developed by MED-EL Medical Electronics and is intended to perform continuous measurements during the insertion of the electrode array in CI surgery, such as Impedance and Field Telemetry (IFT) and evoked Compound Action Potential (eCAP). Figure 1 shows the IMT research software Graphical User Interface (GUI), which allows the user to start the measurement (section A), visualise the IFT measurement (sections D and E), determine manually whether an electrode is inside or outside the cochlea (section B and C) and interrupt the measurements if needed (section F).

The IFT and eCAP values are continuously measured and stored in a local database. In the GUI, measured impedance values are displayed in real time for all 12 contacts (Section C) and updated sequentially in cycles lasting about 1.7 s. Previously measured impedance values are also displayed, in the form of twelve impedance trough time plots (section E).

The measurement setup used for the study is shown in Fig. 2. A Windows-based notebook with the IMT research software, an MAX box, and an MAX coil are used to record the electrophysiological measurements during the insertion of the electrode array. In addition, a video output of the microscope is connected to the computer via an HDMI-to-USB converter (Atomos Connect 4 K HDMI-USB Converter). Video recording software (OBS Software, version 26.1.1.) is used to perform a simultaneous recording of the IMT research software and the microscope output.

The surgical technique used was the same in all patients. After pre-operative setup including full anaesthesia plus regional anaesthesia, facial nerve monitoring (Neurosign 100, Neurosign surgical, Whitland, UK) was installed. Via retroauricular incision, the planum mastoideum was visualised and the periosteum was cut into a U-shape and lifted from the mastoid bone. After identifying the linea temporalis and spina suprameatal, a conventional mastoidectomy was performed. An antrotomy was done and the short process of the incus was visualised. The posterior tympanotomy was performed while continuously monitoring the facial nerve. Then, a bone housing was prepared for the implant and a Synchrony II implant was placed, followed by the opening of the round window. By this time, the IMT research software and measurement setup had been already set up as described in Fig. 2. The measurement was started before coupling the MAX coil to the implant, so the correct coupling could be verified, while confirming the data was being measured and displayed. Once the coupling was successful, the surgeons started the insertion of the electrode array. While the electrode was being inserted in the cochlea, the surgeon provided the operator of the IMT research software with feedback on when each electrode contact was being inserted. Once the insertion was completed, the surgeon informed the user of the IMT research software, and the measurement was stopped. Using the MAESTRO software, routine IFT, eCAP, and eSRT were performed, after which the wound was closed.

Post-operative CT-Scans were conducted to confirm the position of the electrode array.

The first step of the study involves assessing the feasibility of the IMT research software. The feasibility analysis consisted of two parts:

The second step of the study focused on the performance of two methods to determine, in real time, the status of an electrode contact, while the electrode array is being inserted into the cochlea. The status of a contact can either be inside or outside the cochlea.

Two different approaches are used to determine the status of each electrode contact:

Method A is based on continuous impedance measurements (impedance-based), whereas Method B uses the assessment of the surgeon during the insertion in real time (surgeon’s verbal feedback).

In addition, a third method, dubbed “video analysis”, was used to determine the insertion time of each contact. This method was used as a ground truth and is referred to as Method C. The usability of contacts, as determined in the feasibility assessment, is also considered in the performance evaluation. For Method A, only the usable contacts are used for the performance evaluation, while, for Method B, the IMT research software will record the surgeon’s feedback for all contacts.

The goal is to compare the performance of the newly developed method (Method A) and the commonly used method (Method B) against the ground truth (Method C).

Method A consisted in a retrospective analysis of the usable contacts to determine the insertion time. For each contact, its status was marked for each sample, taking the following requirements into account:

Figure 3 shows an example of Method A for case IK_19. It shows the impedance measures over time for each electrode contact. For contacts 6, 7, and 10, it was not possible to determine the status of the contacts from the respective IFT data; therefore, those contacts are deemed unusable and are marked grey. In the usable contacts, the status of a contact is marked red for outside the cochlea, and green for inside the cochlea.

Method B used the surgeon’s feedback during the insertion of the electrode array to determine when each electrode contact was inserted into the cochlea. Section B of Fig. 1 shows how the IMT research software was used to store the feedback provided by the surgeon.

Method C (ground truth) consisted in a retrospective analysis of the video recordings, as obtained from the setup described in Fig. 2 Insertion times for all visible contacts have been determined from the video recordings using the annotation software ELAN [16] which makes it possible to create and visualise annotations synchronised with the video timeline. The insertion times were defined as the last time each contact is seen entering the cochlea, consistently with the definition used in Method A.

In addition to insertion of contacts 1–12, the following events were annotated:

By default, annotations in ELAN are linked to time-intervals of media files and are therefore characterised by both “Begin Time” and “Duration” (or “End Time”) of the annotation. In this work, we chose to align “Begin Time” of annotations with actual insertion times, and to assign arbitrary “Duration” values, which allow for a clear visualisation of the annotations inside the ELAN software. The “Duration” of annotations was only used for visualisation purposes inside the ELAN software and did not represent any value relevant to the analysis.

The performance of the impedance-based method (Method A) and the surgeon’s feedback (Method B) are analysed by comparing them to the ground truth (Method C):

The performance analysis is conducted from ten seconds before the start of the electrode tip insertion until 10 s after the stopper of the electrode array reaches the round window or the surgeon stops the insertion process.

Table 2 presents the feasibility results of using the IMT research software for each case, indicating whether the tool was successfully set up and whether IFT and eCAP measurements were performed. In 19 out of 20 cases, the tool was successfully set up and IFT and eCAP measurements were performed, indicating a feasibility rate of 95%. In 1 out of 20 cases, the setup of the system failed during the coupling of the MAX coil to the implant, which prevented the measurement of IFT and eCAP.

Figure 5A illustrates the number of usable contacts for Method A in the 19 cases where the system was successfully set up and IFT and eCAP measurements were performed. In all cases, a range of 2–12 contacts were found to be usable, with an average of 7.9 usable contacts per case. In contrast, as discussed in the performance evaluation section, Method B recorded the surgeon’s feedback for all inserted electrodes, resulting in 12 usable contacts for each case.

Figure 5B illustrates the occurrence of each contact being usable. Each contact was usable in between 9 and 18 cases out of 19, with an average usability of 12.5 cases (66%). However, contacts 11 and 12 exhibited a higher frequency compared to the other contacts. They were usable in 16 out of 19 cases (84%) and 18 out of 19 cases (95%), respectively.