Introduction
A cochlear implant (CI) is a neuroprosthesis containing an electrode array (EA), which is implanted or introduced into the human inner ear (cochlea) to electrically stimulate the auditory nerve for auditory rehabilitation of patients with severe to profound sensorineural hearing loss. The most crucial step of the whole surgery—the implantation of the EA into the cochlea—is characterized by the mechanical interaction of the “foreign body” with the surrounding intracochlear tissues. Inserting the “foreign body” into the intracochlear biological environment may cause injuries to the delicate soft tissue structures, with the basilar membrane being the most critical as it holds the sensory cells of hearing [
1]. Therefore, traumatic insertions can result in loss of residual hearing [
2]. Moreover, preservation of residual hearing is highly beneficial as many investigations have described improved hearing outcomes [
3,
4] when combining electric stimulation (ES) using the cochlear implant with the natural, residual acoustic (A) hearing. To date, this strategy is widely known as electric acoustic stimulation (EAS) [
5‐
7].
Motivated by these findings, prevention of intracochlear trauma became a dominant topic in cochlear implantation surgery since the 1990s. In order to reduce intracochlear trauma and improve hearing outcomes, Lehnhardt introduced the concept of the “soft surgery technique” already in 1993 [
8]. However, reliable prevention of intracochlear trauma remains a clinical challenge, as the damage resulting from the EA insertion process is influenced by multiple, interdependent factors that are difficult to control. Furthermore, their direct effect on trauma and functional outcome remains difficult to be fully understood, as these factors do not occur isolated from each other. These include the surgical approach (e.g., angle of insertion trajectory, size and location of the opening into the inner ear, level of surgical skills), the type of EA (e.g., stiffness properties, cross-sectional area, surface smoothness), as well as the insertion technique (e.g., insertion velocity). Automation of the insertion process could minimize intracochlear trauma as it allows for:
(a)
standardization, independent of EA characteristics or surgical expertise
(b)
smoothing and optimization, eliminating human tremor and jitter [
9,
10]
(c)
deceleration, enabling ultra-slow continuous insertion velocities beyond those which are humanly feasible [
11,
12].
Initial works on automation of EA insertions emerged in the late 1990s and early 2000s in laboratory settings and consisted on motor-driven insertions of EA prototypes that sought to standardize electrode insertion trials and enhance comparability and reproducibility of results [
13‐
15]. This feed-forward automation was later integrated to robotic insertions of custom-made steerable electrodes [
16,
17]. In those works, during the insertion process the EA is accurately controlled by actuation to facilitate active adaption of the EA’s shape to the patient-specific, spiral-shaped lumen of the inner ear.
Later, the concept of an automated insertion tool to use with standard electrodes in the context of CI surgery was published by Hussong et al. [
18]. However, a clinical application was not conducted, leaving out a realistic description of its sterile use. Our further research was focused on implementation of force sensing capabilities [
19,
20] leading to an increase in complexity due to the integration of more sensitive electronical components. Therefore, development of a solution for sterile use of these tools was disregarded.
More recently, other groups have also developed motorized insertion tools in laboratory experiments [
21‐
23] and although they were also not designed for use under sterile conditions, valuable findings have resulted from these works [
9,
11,
21,
24‐
29]. For example, automated EA insertions have shown its capability in reducing critical peak insertion forces [
9,
10,
30]—subsequently promising to reduce insertion trauma. Different insertion velocities have been explored using automated setups, and to date, most data suggest that lower insertion velocities produce lower insertion forces [
11]. Moreover, these automated insertions can be programmed to occur at speeds slower than 0.9 mm/s, which according to the work of Kesler et al. [
12] are not manually feasible as a continuous, steady movement.
Recently, two approaches for robotic surgical systems developed for clinical-, intraoperative use have been described. The robotics-assisted surgical tool proposed by Kaufmann et al. has already published promising results in its pre-clinical validation phase [
10], but has the drawback of requiring additional drilling and screwing to the patient’s head to fixate the tool. In addition, a teleoperated robot (RobOtol, Collin Orl, Bagneux, France) has been used by Nguyen et al. to perform EA insertions in first patients [
31]. However, this system requires the use of a robotic arm, which may significantly increase CI surgery costs and potentially limit patient access to this technology.
As an alternative, we introduce a new concept of a simple tool designed to automate the feed-forward motion of EAs or intracochlear catheters in CI surgery. Its design aims to achieve a maximum simplicity to facilitate wide clinical and surgical translation while still complying with surgical necessities (e.g., sterility).
The scope of this manuscript is to (1) present the ideation and design details of the tool, herein after referred to as Cochlea Hydro Drive (CHD), and (2) test the defined main features such as automation capabilities, handling and adaptation to a surgical-like scenario.
Discussion
The here presented prototype fulfills conformity with all needs for sterility. This is achieved by a strong simplification of the concept and the reduction to the essential: use of a standard infusion pump, a disposable commercially available syringe and a few small size adapters. By repurposing a syringe as a hydraulic cylinder, one can avoid sterilization or sterile housing of an electrically powered actuator, such as stick–slip piezo drives [
18], or rotary motor [
21] as previously reported. All components for energy generation and motion control are located outside the sterile area of the OR as they are integrated in the infusion pump. Furthermore, standard infusion pumps are already optimized for providing very low but also very precise flow rate, which can be adjusted in a wide range.
Our initial testing shows that repurposing a syringe as a hydraulic cylinder facilitates hydraulic actuation. A feed-forward motion without significant peaks or breaks was observed for all three programmed velocities. We observed deviations in the resulting average velocity compared to the theoretical calculated values. This is better appreciated in the trials corresponding to 0.11 mm/s, where the resulting velocity was 0.07 mm/s. However, this finding does not contradict the general concept as these results still fall within the desired range of “ultra-slow” velocities. Additionally, accuracy in setting the velocity of the plunger can be improved by further testing and calibration of the system and re-evaluation of the workflow.
Insertions of an EA into an aCM confirmed that our tool is indeed able to hold and insert an EA in response to hydraulic actuation. These trials showed smooth force profiles, especially with an ultra-slow velocity of 0.03 mm/s, which in turn resulted in mean maximum forces of 0.060 ± 0.007 mN (Fig.
4a). The mean insertion forces resulting from our experiments are comparable to previous reports using MED-EL electrodes [
36,
37]. However, caution is warranted when comparing our tool’s absolute force values with other studies, as different methodologies pertaining to the cochlea model (geometry, material, lubricant), EA and insertion depth may also impact the resulting insertion forces. Therefore, the most valuable finding from the present insertion experiments using the CHD is that the resulting insertion profiles do not show numerous peaks, pauses or breaks, in contrast to manually performed insertions, as previously reported in [
9,
12,
30].
In case of leakage in the tubing or connectors only a part of the fluid flow will reach the CHD which causes a decrease in the insertion velocity (or even a full stop of the movement depending on the size of the leak). This may be an argument to use sterile or saline water instead of regular water in the operating room, therefore avoiding an additional risk for the patient.
Incorporation of the 3-way stopcock valve allows for sufficient venting of the tubing. When that few small air bubbles remained inside the valve, no impact on the movement of the device was observed. Furthermore, this allows for an immediate cessation of feed-forward motion of the CHD when used to stop the insertion.
In its current design, automated backward motion after implantation is not possible. The plunger needs to be moved backwards (inside the syringe of the CHD) manually or indirectly by removing the larger syringe (of the syringe driver) and pulling its plunger in order to generate a negative pressure. The 5 ml syringe allows for a travel range of approx. 45 mm which covers full insertion depth of even the longest EA in the market (with 31.5 mm). Until now, our tool was developed for straight (lateral-wall) EAs.
A drawback of the simplicity of the CHD is that a force sensing capability is not integrated as in previous tools [
19,
20]. Therefore, monitoring of the insertion process based on insertion forces is not possible in the current design of the CHD. However, we consider this as an acceptable limitation, since the current version of the tool is designed to facilitate automation of the EA insertion and potentially allow for very slow insertion speeds and therefore reduced insertion forces [
11]. In addition, the current design of the CHD allows for its use with the conventional mastoidectomy with a facial recess approach, which is an “open” surgery access to the inner ear. Thereby, the surgeon can still visually observe the insertion process and stop the automated feed of the implant if irregularities (e.g., EA buckling) occur. One has to investigate in further studies whether the loss of the surgeon’s capability to manually sense at least some insertion forces [
38] is a critical issue; especially as the soft surgery protocol suggests cessation of the insertion when resistance is met. Additional investigations are required to elucidate whether visual observation of the insertion process ensures the same safety for the residual hearing as the haptic feedback. This is of particular importance in patients with functional residual hearing.
The implementation of a tool like the CHD could bring other benefits. For example, the device could later on be combined with additional intraoperative measurements such as fluoroscopy [
39], or cochlear monitoring [
40] in order to achieve a deeper understanding of the electrode insertion process and the underlying mechanism of intracochlear trauma. Ultimately, more information and strategies could be gathered to develop different approaches that could guarantee preservation of residual hearing.
In its current version, the tool is intended to be positioned using a flexible arm attached to a standard surgical retractor. This requires sufficient experience of the surgeon in estimating the best trajectory into the basal turn of the cochlea for the individual patient [
41]. However, this is also true in conventional, manually performed electrode insertion and therefore not a specific challenge when using the CHD. In addition, using the surgical retractor, automation of probe insertions goes without additional invasive steps, such as screwing into the skull, as required in the case of the iotaMotion system in order to drive the unit [
10].
In case of using a tool like the CHD for EA insertion, the alignment to an individually planned trajectory can be improved by incorporating image-guided surgery systems [
22,
42,
43] or micro-stereotactic frames [
44,
45] into the surgical setting. In doing so, an optimized trajectory can be planned based on patient-specific imaging and followed in the OR with an accuracy outperforming what is manually feasible. This is another possibility that can be further explored in the future.
The limitations of this work include that we did not perform probe insertions into human cadaveric cochlear specimens with the use of the CHD. However, we have explored our initial concept to justify moving forward with such experiments. Also, the learning curve of assembling, handling and positioning of the tool is herein not fully characterized, but will be further explored. Likewise, the variability on the resulting velocities at which the CHD response needs to further validated and replicated.
To summarize, more detailed investigations have to address questions such as: How accurately can the device be positioned manually along an appropriate insertion axis? Is there a steep learning curve or how does these results depend on the experience of the surgeons? What about the inter-operator variability when mounting, positioning and using the CHD? How reliably can the actual insertion of the EAs into the cochlea be performed using the tool? How do different velocities impact insertion trauma? However, answering these questions is reserved for further studies.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.