Robotic Versus Manual Electrode Insertion in Cochlear Implant Surgery: An Experimental Study

Article information

Clin Exp Otorhinolaryngol. 2025;18(1):21-29

Publication date (electronic) : 2024 December 23

doi : https://doi.org/10.21053/ceo.2024.00253

Salman F Alhabib^,1,2, Farid Alzhrani^1,2, Abdulrahman Alsanosi^1,2, Mariam Al-Amro^1,2, Abdulaziz Alballaa^1,2, Ibrahim Shami³, Abdulrahman Hagr^1,2, Asma Alahmadi^1,2, Tahir Sharif⁴, Maximilian Stichling⁵, Marco Matulic⁶, Masoud Zoka Assadi⁵, Yassin Abdelsamad⁴, Fida Almuhawas^1,2

¹King Abdullah Ear Specialist Center (KAESC), King Saud University Medical City, Riyadh, Saudi Arabia

²Department of Otolaryngology-Head and Neck Surgery, College of Medicine, King Saud University, Riyadh, Saudi Arabia

³King Fahad Medical City, Ministry of Health, Riyadh, Saudi Arabia

⁴Research Department, MED-EL GmbH, Riyadh, Saudi Arabia

⁵R&D, MED-EL GmbH, Innsbruck, Austria

⁶CASCINATION AG, Bern, Switzerland

Corresponding author: Salman F Alhabib King Abdullah Ear Specialist Center (KAESC), King Saud University Medical City, King Saud University, Riyadh 11411, Saudi Arabia Tel: +966-11-477-4136, Fax: +966-11-477-5524 Email: sfalhabib@hotmail.com

Received 2024 August 23; Revised 2024 December 8; Accepted 2024 December 21.

Abstract

INTRODUCTION

Cochlear implants (CIs) have revolutionized the treatment of profound deafness by enabling individuals with significant auditory deficits to hear through direct stimulation of the auditory nerve [1,2]. Recently, eligibility criteria for CIs have expanded to include patients with residual hearing. This expansion makes the precision of electrode placement critical for optimizing auditory outcomes. Achieving this precision requires the slow and consistent insertion of the electrode array within the scala tympani to reach a specific cochlear coverage target.

Studies have consistently shown that insertion speeds of 0.25 mm/sec or slower significantly reduce intra-cochlear force and pressure, thereby preserving hearing and maintaining the integrity of the vestibular system [3]. Pre-surgical planning for cochlear implantation, which incorporates both audiological and anatomical information, is essential for determining the precise insertion depth of the electrode. This ensures optimal cochlear coverage and frequency response for the patient. Optimal placement can be achieved by either fully inserting an electrode of the appropriate length or by partially inserting a longer electrode.

Proper electrode positioning is crucial to prevent complications such as tip fold-over and scala translocation. Tip fold-over occurs when the electrode tip bends back on itself, which can hinder effective stimulation of the auditory nerve and potentially damage the cochlear structures. Scala translocation involves the unintended movement of the electrode from the scala tympani to the scala vestibuli, leading to diminished hearing outcomes and increased cochlear trauma. Correct placement of electrodes within the scala tympani is vital for enhancing auditory rehabilitation and reducing the risk of these adverse events, thereby contributing to the long-term success and safety of CI operations [4,5].

The inherent complexity of inner ear anatomy, along with the limitations of manual surgical dexterity, raises concerns about consistently achieving precise electrode insertion. Previous studies have shown that maintaining a continuous insertion speed of 0.25 mm/sec for CI electrodes is beyond human capability; the minimum continuous forward insertion speed achievable using manual techniques averages 0.87 mm/sec [6]. Despite advancements, significant variability in hearing preservation outcomes persists. This variability is influenced by factors such as surgical expertise, patient demographics, implant centers, and electrode designs, leading to many patients experiencing a decline in their residual hearing over time. Over the past few decades, numerous research groups have endeavored to create robotics and motorized solutions to tackle specific challenges. Despite numerous innovations, the majority of these projects have not yet advanced to the clinical stage [7-10]. Recently, two commercial systems, the iotaSOFT Insertion System (iotaMotion Inc.) and RobOtol (Collin), have reached commercialization and are currently being introduced clinically. The availability of these commercial systems marks a significant step forward in achieving precise electrode insertion [11,12]. However, the primary challenge that remains is to demonstrate clinical evidence supporting the advantages of robotic systems over traditional manual methods.

Most recently, OTODRIVE (CASCINATION AG) has been added to the portfolio of otological robotic systems. This system allows for motorized motion at user-defined speeds, which can be employed for robotic electrode array insertion [13,14]. The experimental study compared the quality and precision of manual electrode insertion to that of robotic (OTODRIVE-assisted) electrode insertion.

MATERIALS AND METHODS

Device description

In this study, five devices were utilized to enable robotic electrode insertion: the OTODRIVE (CASCINATION AG), Forceps OD, Connector OD, OTOARM Aligner, and OTOARM (MEDEL), as illustrated in Fig. 1.

Fig. 1.

Schematic overview of the set-up, including the OTODRIVE, Forceps OD, Connector OD, OTOARM Aligner, and OTOARM.

Study setting and protocol

This experimental study was conducted at the temporal lab of a tertiary CI center. It was performed on formalin-fixed full cadaveric heads (n=10, 10 right and 10 left temporal bones) with approval from the Institutional Review Board of King Saud University, College of Medicine (No. E-24-8765). The requirement for informed consent was waived due to the retrospective nature of this study. For each head, senior neurotologists (n=9 surgeons) performed manual and robotic insertions.

Preoperative imaging

A computed tomography (CT) image with 0.4 mm resolution was acquired from all the samples.

Preoperative analysis

Three different electrode lengths, FLEX26 (n=4), FLEX28 (n=8), and FLEXSOFT (n=8), were used in this study (MED-EL). The full electrode lengths for FLEX26, FLEX28, and FLEXSOFT are 26 mm, 28 mm, and 31 mm, respectively, and all electrodes feature 12 contacts (C1-C12). Using OTOPLAN software V5 3.1.0 (CASCINATION AG), each temporal bone’s cochlear duct length (CDL; mm) was measured. For each electrode, the estimated C1 (most apical contact) insertion angle (degrees) and cochlear coverage (%) were calculated. The electrode selection for each temporal bone ensured a cochlear coverage of more than 75%. Cochlear coverage was defined as the percentage of the CDL covered by the most apical contact (C1) of the inserted electrode array. Full electrode insertion was planned for all cases.

Training

The day before the experiment, the surgeons practiced the robotic insertion on a plastic phantom and reviewed the instructions for use.

Sample preparation

The heads for CI insertion were prepared the day before, following the usual steps for mastoidectomy with facial recess access. A round window niche was drilled, and maximum exposure of the round window was achieved. No cochleostomy or extended round window was required in any of the samples. No special adjustments were made in preparation for this study.

Manual insertion

The manual insertions were performed on the left side with the selected electrode. Microscopic images of the insertions were recorded for post-processing of the insertion speed.

Robotic insertion

(1) Set-up: the OTOARM was clamped to the operating table rail on the contralateral side of the insertion. The OTOARM Aligner was mounted on the OTOARM. Next, the OTODRIVE handpiece was attached to the OTOARM Aligner, and the compatible sterile drape was placed. Finally, the Connector OD and Forceps OD (MED-EL) were attached to the tip of the OTODRIVE handpiece. (2) Robotic insertions: initially, the Forceps OD was driven to the most distal position of the handpiece tip (χ=40 mm). Then, the surgeon would use the OTOARM and OTOARM Aligner to position and orient the tip of the Forceps OD to achieve the desired insertion trajectory. Once the trajectory was set, the Forceps OD was driven back to the most proximal position (χ=0 mm). Then, the electrode was mounted on the Forceps OD on the lead, approximately 5 mm above the stopper. The surgeon then inserted the electrode by pressing the right foot pedal. All the robotic insertions were performed at 0.1 mm/sec for the whole insertion. If required, the surgeon uses assistive tools such as a surgical claw to assist with the insertion process. The insertion was stopped once the full length of the electrode was achieved or if further insertion was not possible due to buckling. After the insertion, the electrode was removed from the Forceps OD, and the OTOARM was moved away from the surgical field. Fig. 2 shows the robotic insertion setup (Fig. 2A) and the electrode was then fixed at the round window (Fig. 2B).

Fig. 2.

(A) The robotic insertion setup. (B) The electrode is inserted through the round window using OTODRIVE.

Postoperative imaging

A postoperative CT image with 0.4 mm resolution was acquired from all the samples.

Postoperative analysis

The postoperative analysis was conducted using the OTOPLAN software (version 3.1.0, V5). This process automatically registers the postoperative CT with the preoperative CT scan to assess electrode placement and insertion status. In the first step, the preoperative image, previously used for analysis and containing a reconstructed three-dimensional (3D) model of the inner ear and cochlear scalae, was loaded into the software. Subsequently, the postoperative image was imported and fused with the preoperative image and model using a mutual information algorithm. Next, the software automatically detected and reconstructed the electrode array in 3D, calculating the insertion angle (degrees) and cochlear coverage (%). Finally, the electrode’s location within the scala tympani and any instances of tip fold-over or scala deviation were visually evaluated by overlaying the electrode array on the cochlear scalae 3D model. The following metrics were also evaluated during the analysis.

(1) Electrode coverage (%): the electrode coverage percentage was calculated by dividing the inserted length of the electrode by the full electrode length. (2) Suboptimal insertion: suboptimal insertion refers to cases with incomplete insertion, where at least one electrode contact was outside the cochlea or over-insertion of the electrode when the stopper of the electrode was inserted inside the cochlea. (3) Insertion complications: the presence of tip fold-over and scala deviation were recorded for each insertion. Additionally, for robotic insertions, any cases where the surgeon could not insert the electrode due to issues to the following reasons: 1) inability to set-up the devices, 2) blocked view of the surgical field by the Forceps OD or OTODRIVE handpiece, 3) inability to set the desired insertion trajectory by the OTOARM and OTOARM Aligner, 4) insufficient range of motion, or 5) complications with using the assistive tools such as the surgical claw—were also recorded as insertion complications.

Manual insertion speed

The average insertion speed of the manual insertions was evaluated retrospectively, using recorded microscopic images at eight frames per second. First, the “total insertion time (second)” and the “pause durations (second),” where the insertion was interrupted (e.g., stops or electrode repositioning), were recorded for each case. The average insertion speed was then calculated by dividing the inserted length of the electrode by the movement duration, which is the total insertion time minus the pause durations:

Manual insertion speed (mm second )= inserted length of electrode (mm) total insertion time (second)-pause duration (second)

RESULTS

In total, 10 robotic and 10 manual insertions were performed on the right (R) and left (L) ears. Following each insertion, a postoperative CT scan was acquired, and the resulting images were analyzed. An example of this postoperative analysis, case 7L, is presented in Fig. 3. All insertion parameters were analyzed using 3D reconstructions of the full cochlea (red), scala tympani (purple), and the electrode array with contacts C1 to C12 in relation to the scala tympani. Table 1 displays the preoperative and postoperative cochlear and electrode insertion parameters. These include CDL (mm), the selected electrode for each temporal bone, corresponding pre- and post-insertion angles (degrees), pre- and post-cochlear coverage (%), postoperative insertion depth (mm), and electrode coverage (%).

Fig. 3.

Postoperative analysis of case 7L using the OTOPLAN software. Insertion status of the electrode array inside the three-dimensional reconstruction of the full cochlea in red (A) and scala tympani in purple (B). CC, center of cochlea; CRW, center of the round window; L, left.

Table 1.

Preoperative and postoperative cochlea and electrode insertion parameters, including CDL, the selected electrode for each temporal bone, corresponding pre- and post-insertion angles, pre- and post-cochlear coverage, postoperative insertion depth, and electrode coverage

Table 2 presents the evaluated means±standard deviations for insertion angle, cochlear coverage, electrode coverage, and insertion speed. For all three measurements, the P-values exceeded 0.05 (Mann-Whitney U-test), suggesting that there was no statistically significant difference between the manual and robotic insertion techniques regarding insertion angle, cochlear coverage, or electrode coverage (Fig. 4).

Table 2.

Postoperative insertion angle, cochlear coverage, and electrode coverage between the manual and robotic techniques

Fig. 4.

Postoperative insertion angle (A), cochlear coverage (B), and electrode coverage (C) between the manual and robotic techniques. The solid gray line represents the mean for each group.

Four cases of incomplete insertion were observed during robotic insertions. Specifically, two cases (4R and 10R) involved C10, one case (5R) involved C11, and one case (6R) involved C12, all located at the round window membrane. There were no instances of over-insertion, basal buckling, tip fold-over, scala deviation, or any other insertion complications associated with robotic techniques. Conversely, manual insertions exhibited two cases of incomplete insertion: one case (5L) with C9 and another case (1L) with C10, both at the round window membrane. Additionally, as Fig. 5 shows, there was one case of over-insertion (3L) and two cases of partial insertion (cases 1L and 5L), and one case of basal buckling (case 4L). Unlike the robotic insertions, no instances of tip fold-over or scala deviation were noted in the manual insertions.

Fig. 5.

Preoperative estimation and postoperative actual C1 insertion angle for manual (A) and robotic (B) insertions. The cases with insertion complications are marked by a specific symbol. L, left; R, right.

A representation of preoperative estimation and postoperative C1 insertion angle, along with observed suboptimal insertions and complications, is presented in Fig. 5 on a case-by-case basis. In cases 2L, 1R, 8R, and 2R, the postoperative insertion angle exceeded the preoperative estimation, attributed to errors in predicting the preoperative insertion depth as described in the literature [15-17]. However, the electrode’s stopper was positioned before the round window membrane in all these cases, indicating that over-insertion did not occur.

The insertion speeds for manual and robotic techniques were, on average, 0.66±0.31 mm/sec and 0.1 mm/sec, respectively. A statistically significant difference in insertion speed between the manual and robotic techniques was observed (t-test, P<0.001). Table 3 displays the inserted length, movement duration, pause duration, and calculated average insertion speed for manual insertions. For all robotic insertion cases, a consistent speed of 0.1 mm/sec was maintained.

Table 3.

Inserted length, movement duration, pause duration, and calculated average insertion speed for manual insertion cases

DISCUSSION

This study compared the efficacy and precision of manual versus robotic (OTODRIVE-assisted) electrode insertions in cochlear implantation. Both manual and robotic techniques demonstrated similar performance in achieving the desired insertion angle, cochlear coverage, and electrode placement. This suggests that robotic assistance maintains the precision of electrode placement, despite being a relatively new approach for the surgeons involved. The insertion speed for robotic insertion was 0.1 mm/sec, significantly slower than the manual insertion, which had a mean speed of 0.66±0.31 mm/sec. This controlled speed is crucial, as previous studies have indicated that slower insertion speeds can reduce intra-cochlear force and pressure, potentially helping to preserve residual hearing and vestibular function. A study by Rajan et al. [18] showed that a slow electrode insertion speed (0.25 mm/sec) facilitated complete electrode insertion, reduced the occurrence of insertion resistance, and helped preserve residual hearing and vestibular function after cochlear implantation. The average insertion speed calculated in this study is consistent with the findings of Kontorinis et al. [19], where retrospective measurements of insertion speed through microscopic videos of 116 human implantations indicated speeds ranging from 0.7 to 2.75 mm/sec.

In this study, robotic insertions did not lead to any complications such as tip fold-over, scala deviation, or issues like setup difficulties, obstructed views of the surgical field by the Forceps OD or OTODRIVE handpiece, challenges in setting the desired insertion trajectory with the OTOARM and OTOARM Aligner, limited range of motion, or problems with using assistive tools like the surgical claw. However, there were four instances of incomplete insertion. Conversely, manual insertions resulted in two cases of incomplete insertion, one case of over-insertion, and two instances of basal buckling.

In case 5, both manual and robotic insertions resulted in incomplete insertion. In case 4, while the robotic method also led to incomplete insertion, manual insertion caused basal buckling, indicating that the robotic insertion did not apply enough force to cause this complication. In case 6R, which was categorized as an incomplete robotic insertion, C12 was positioned at the round window membrane, making it still suitable for fitting. The average insertion angle for robotic insertion in this study was 506°± 103°, significantly higher than the 321°±84° reported in a previous study by Kaufmann et al. [20], which also compared manual and robotic insertions. This increase is attributed to the use of longer electrodes in the current study. This study utilized defrosted formalin-fixed cadaveric heads, which may not perfectly mimic the soft tissue properties of the cochlea that are prone to tissue deformations. This could have influenced the incidence of incomplete insertions observed in both manual and robotic methods due to the rigidity of the cadaveric tissues. This observation aligns with findings from previous studies on cadaveric samples [21,22].

In terms of expertise, the study involved senior neurotologists performing the insertions. The results may vary with less experienced surgeons, and further studies could explore the impact of surgeon expertise on the outcomes of robotic-assisted insertions. The OTODRIVE robotic system, as previously reported, offers a consistent and controlled insertion process that could be particularly beneficial in standardizing CI operations across various surgeons and clinical settings. This consistency may help reduce the variability in hearing preservation outcomes that is commonly observed with manual insertions. Although the controlled and slower insertion speeds achievable with robotic systems might theoretically lessen the risk of intra-cochlear trauma and improve the preservation of residual hearing, these expectations have yet to be fully validated through larger clinical studies. Current findings underscore the importance of investigating the role of robotic systems in enhancing the precision of electrode placement, which is crucial for achieving optimal auditory outcomes.

Although recent evidence suggests clinical benefits of using robotic insertion systems [23-25], further research is necessary to confirm whether these systems consistently outperform manual techniques in terms of long-term auditory outcomes and hearing preservation. This research should involve fresh samples that are well-defrosted ahead of time, use Thiel fixation, and/or include larger patient cohorts. Additionally, studies should explore the variability in patient outcomes and the practical challenges of robotic-assisted surgery. These challenges include the additional time and training required for surgeons, given the learning curve associated with such systems. Moreover, setting up robotic equipment can be more time-consuming than manual methods, potentially affecting the efficiency of operating room workflows. Another significant limitation is the absence of haptic feedback in robotic systems, which prevents surgeons from feeling the resistance during electrode insertion, unlike with manual tools. A robust body of evidence is essential to inform clinical decision-making and assess the broader implications of integrating robotic systems into routine cochlear implantation operations.

This study demonstrated that robotic-assisted electrode insertion is comparable to manual insertion in terms of precision. However, it offers significant advantages in controlling the insertion speed, which may reduce intra-cochlear pressure changes during the procedure. These findings support the continued development and clinical evaluation of robotic systems for CI surgery, highlighting the importance of slow and controlled electrode insertion to preserve hearing and improve surgical outcomes. Further research is necessary to validate these findings in clinical settings and to investigate the long-term benefits of robotic-assisted cochlear implantations.

HIGHLIGHTS

Article information Continued