Redefining Cochlear Implant Strategies: Auditory Nerve Excitability in Scala Vestibuli vs. Scala Tympani Placements

Significance 

Hearing loss is a major health issue that affects millions of people around the world. It does more than just make conversations difficult—it can impact cognitive function, interfere with work and social interactions, and lower overall quality of life. For those with severe or complete hearing loss, cochlear implants (CIs) have been life-changing. These devices work by bypassing the damaged hair cells inside the cochlea and sending electrical signals directly to the auditory nerve which restores a sense of hearing. However, CIs have some limitations, especially for patients with structural abnormalities or blockages inside the cochlea. When standard electrode placement is not an option, alternative surgical techniques need to be explored to ensure the implant can still effectively stimulate the auditory nerve. The most common approach for cochlear implantation involves placing the electrode array in the scala tympani (ST), which is the lower duct of the cochlea and this location is preferred because it is easier for surgeons to access and sits close to the spiral ganglion neurons that relay sound signals to the brain. But not everyone has a clear path for ST electrode placement. Some patients develop ossification, fibrosis, otosclerosis, or cochlear obstructions due to meningitis or congenital abnormalities, all of which can make it impossible to insert the electrode properly. For these cases, an alternative option is to place the electrode in the scala vestibuli (SV), the upper duct of the cochlea. While SV placement has been used in some patients, it is not yet well understood how this approach compares to the standard ST placement in terms of hearing outcomes.

A major question surrounding SV implantation is whether it stimulates the auditory nerve as effectively as ST placement. For a CI to work well, it needs to precisely activate specific nerve fibers while avoiding unnecessary stimulation of neighboring fibers. This ensures that the brain interprets electrical signals clearly, allowing the person to recognize different sounds and speech patterns. However, several factors—such as differences in how electrical currents spread through cochlear fluids, variations in nerve fiber positioning, and the structural properties of surrounding tissues—could affect how well the SV electrode stimulates the nerve. To this account, a new research paper has been published in Journal of Neural Engineering authored by Dr. Andreas Fellner, Dr. Cornelia Wenger, and Professor Frank Rattay from the Vienna University of Technology alongside Dr. Amirreza Heshmat from University of Texas MD Anderson Cancer Center. They built a detailed three-dimensional finite element model (FEM) of the human cochlea, reconstructed from high-resolution micro-CT scans, to replicate the structure and electrical properties of the cochlea as accurately as possible in order to analyse how auditory nerve fibers respond to different electrode placements.   

The research team used monophasic anodic and cathodic pulses, each lasting 50 microseconds to compare nerve excitability between ST and SV placements. They were able to map out the extracellular voltage distributions along the nerve fibers using FEM calculations. These results were then fed into a multi-compartment neuron model, which helped predict where and how action potentials (APs) were triggered. Essentially, they were able to determine the minimum electrical charge needed to stimulate the nerves (threshold currents) and see how different parts of the cochlea responded to stimulation. One of the most eye-opening discoveries was the difference in stimulation efficiency between ST and SV electrodes, particularly when looking at pulse polarity. When using cathodic (negative) stimulation, SV electrodes required less current to activate the auditory nerve fibers than ST electrodes. That means SV placement could be an advantage for patients who cannot have an ST implant, since it takes less energy to get the nerve fibers firing. But the opposite happened with anodic (positive) pulses—SV electrodes needed much higher currents compared to ST electrodes. This difference shows that where an electrode is placed in the cochlea significantly impacts how it interacts with nerve fibers, making pulse polarity selection a crucial factor in implant programming. Afterward, the authors looked at how far the stimulation spread. This is a big deal for sound clarity because ideally, an electrode should only activate the intended nerve fibers and not spill over to neighboring ones. Their findings showed that SV electrodes were more selective. On the other hand, ST electrodes had a broader spread, especially when higher currents were used. The authors also found that moving an electrode by as little as 3 to 9 degrees could dramatically change the amount of current needed to trigger nerve activity. In some cases, shifting the electrode even slightly lowered the threshold significantly, while in others, it made stimulation much harder. This shows how important precise electrode placement is during surgery, as even minor adjustments can make a noticeable difference in the excitability of auditory nerve fibers.  

To ensure their model was as realistic as possible, the researchers compared their simulated intracochlear voltage data with real-world clinical measurements taken from actual CI users. Specifically, they looked at voltage decay patterns along implanted electrode arrays and found that their FEM simulations closely matched recorded patient data. This was an important validation step, proving that their model was not just a theoretical exercise, but could accurately reflect what happens inside a human cochlea. Because of this, their findings could have a direct impact on clinical decision-making, particularly when it comes to choosing the best electrode placement strategy for each patient. The team also investigated where the APs actually started in the auditory nerve fibers. In most cases, whether the electrode was in the ST or SV, the APs were triggered at the peripheral dendritic terminal of the nerve. This suggests that this region is the primary target for electrical stimulation by such lateral-wall electrodes, making it a critical area to consider when designing electrode arrays while also bearing the patient and frequency specific neural degeneration status in mind. However, when stimulation intensity was very high, some APs were triggered closer to the center of the nerve, which could reduce the precision of frequency discrimination at stronger stimulation levels. To take their findings a step further, the team also tested suprathreshold stimulation, meaning they increased the current beyond the minimum level needed to activate the nerve. This was an important part of the study because real-world listening environments involve a wide range of sound intensities, and CIs need to be able to adjust accordingly. What they found was that SV electrodes maintained better selectivity at higher currents. Meanwhile, ST electrodes had more widespread activation.

In conclusion, the research from the Vienna University of Technology could change how CIs are used and the new innovation can provide solutions that were once considered impossible. It demonstrated that SV implantation is more than just a backup plan—it is a real and effective alternative. In fact, cathodic stimulation in the SV required lower current thresholds, therefore, less energy is needed to activate the nerves. This has direct implications for both CI programming and surgical planning, allowing doctors to adjust stimulation settings to improve patient outcomes. Another major advantage of the new technology is related to how precisely the electrodes stimulate the nerve fibers, which is critical for clear and natural sound perception because it showed that SV electrodes are more selective than ST electrodes, especially at higher stimulation intensities. Therefore, patients may experience sharper pitch perception which is important in noisy environments where speech clarity is easily lost. Indeed, the results suggest that electrode positioning is more important than previously thought which could lead to new electrode designs that improve sound quality.

Redefining Cochlear Implant Strategies: Auditory Nerve Excitability in Scala Vestibuli vs. Scala Tympani Placements – Advances in Engineering

VIDEO CAPTION: This video shows parts of the finite element model which is used to calculate the induced electric potential along the target auditory nerve fibers (ANF) following electric stimulation by a cochlear implant (CI). The outer cochlear bone is shown in yellow at the beginning of the video. The cochlear ducts, namely the scala tympani (ST), the scala vestibuli (SV) and the scala media, are shown in turquoise color and keep visible throughout the video. A CI consists of a silicon carrier (grey) and active electrode contacts. In this video four different CI locations are presented, with lateral-wall (lw) and peri-modiolar (pm) positions in either the lower ST or upper SV. The model includes 25 realistic ANF pathways as obtained by manual tracing using µCT dataset of a human cochlea. The location of the soma is highlighted by a bigger point. Only one electrode contact (red) at respective locations of the four CIs is presented throughout the video which targets the ANF at an insertion angle of 96°. The induced electric potential (in mV) for all ANFs corresponds to the colorbar at the right. Shortly the induced electric potential along the whole electrode carrier is shown. We have tested 25 different electrode contact positions, one for each ANF. Details of the cochlear and ANF models  can be found in [1].

[1] Fellner, A., Wenger, C., Heshmat, A., and Rattay, F. (2024). Auditory nerve fiber excitability for alternative electrode placement in the obstructed human cochlea: electrode insertion in scala vestibuli versus scala tympani. Journal of Neural Engineering, 21 (046034). https://doi.org/10.1088/1741-2552/ad6597

Redefining Cochlear Implant Strategies: Auditory Nerve Excitability in Scala Vestibuli vs. Scala Tympani Placements - Advances in Engineering
Propagating action potentials (AP) are plotted for threshold stimulation of an intact and healthy ANF96, which is the target fiber presented in the video. The four panels correspond to the different CI locations as indicated in the respective titles including the cathodic threshold. Note that the lwSV array induces the lowest cathodic threshold (bottom right), followed by pmST (top left), and lwST (bottom left), and the highest threshold is registered for the pmSV array (top left). ANF96 is a descriptive example since the average cathodic threshold values of all 25 ANFs follow the same ranking. Each line shows the transmembrane potential over time (x-axis) at one specific compartment of the ANF, whereas the peripheral nerve terminal is plotted in blue, the soma compartment in red, and peripheral and central nodes in black and grey respectively. The vertical distance between the lines corresponds to the geometric distance between compartments. Although the APs are always initiated in the terminal, the excitation behavior in the dendrite is different among CIs with variations in areas of initial hyper – or depolarization. Furthermore, note the differences in spike timing as highlighted by the colored crosses at the time axis (in ms) which indicates the time of maximal transmembrane potential in the terminal and soma. Earlier AP initiation is predicted for the lw CIs which are closer to the terminal where the AP forms. The longest delay is registered for the pmST (initially depolarized terminal) which results in a propagation delay of about 0.6ms in comparison to the fastest lwSV (directly hyperpolarized terminal).

 

Redefining Cochlear Implant Strategies: Auditory Nerve Excitability in Scala Vestibuli vs. Scala Tympani Placements - Advances in Engineering
Absolute thresholds of 25 target ANFs are presented for stimulation with 50 µs long monophasic cathodic pulses emitted by four different CI locations: peri-modiolar ST (pmST) in blue, peri-modiolar SV (pmSV) in red, lateral-wall ST (lwST) in green, lateral-wall SV (lwSV) in orange. In the left panel thresholds for intact ANFs (x-axis) are compared to the thresholds for severely degenerated ANFs (y-axis) with a reduced dendritic diameter (from the default value of 1.3 µm to 0.5 µm). In the right panel results are presented for completely degenerated ANFs that have lost the whole dendrite, with the same y-axis scale for better comparability. Corresponding mean threshold values of each CI are plotted as colored crosses at the respective axis. For intact ANFs the lowest to highest cathodic threshold value is found for lwSV, pmST, lwST, pmSV respectively. For severely and completely degenerated fibers pm and ST CIs show on average lower thresholds due to their proximity to the intact nerve structures.

About the author

Cornelia Wenger received her MSc in Technical Mathematics with concentration in Mathematics in Science from the Vienna University of Technology, where she focused her research on Biomedical Engineering. She started with the standard academic career beginning with a PhD fellowship (neuronal excitation and cochlear implants, completed in 2012) followed by a PostDoc position (non-invasive brain and spinal cord stimulation), before transitioning to industry (cancer treatment with electric fields). A common research area over the years was the development of computational models at different scales (sub-cellular, single cells, small-scale cellular networks, realistic full body, patient cohort models) for numerical simulation of electromagnetic behavior. As a Clinical Research Professional she was involved in developing, validating and certifying new technology for the treatment of solid tumors. She is widely published, has presented her work at various international conferences, and has been an invited lecturer during several academic and clinical forums.

About the author

Andreas Fellner received his MSc in Business Informatics in 2006 from the Vienna University of Technology. After several years working in process automation and optimization in the steel industry, he returned back to the Vienna University of Technology, where he received his MSc in Biomedical Engineering in 2017 and his PhD in 2020.  In his professional and academic career, he has always focused on the modeling and simulation of complex problems and the development of computational frameworks. In the past years, he mainly investigated neuronal response with the finite element method, where he authored several journal and conference articles and presented his work on international platforms.

About the author

Frank Rattay is specialized in computational neurosciences, biophysics, and biomedical engineering. He is a leading expert in the theory of functional electrical nerve and muscle stimulation. In 1986 he introduced the activating function concept, which is the most cited method to explain extracellular electrical excitation. Concerning electrical stimulation of the cochlea, he made the first modeling studies in the field on a hybrid computer (Hochmair-Desoyer et al. 1984). For the second paper (Motz and Rattay 1986) the authors had a discussion on the differences between intra- and extracellular stimulation with Andrew F. Huxley, Nobel prize winner for describing the excitation of the neural cell membrane. The missing knowledge of the scientific community regarding intra- versus extracellular stimulation inspired Rattay to develop the activating function concept (Rattay 1986). Because of such lacks of knowledge Rattay wrote the first monograph ‘Electrical nerve stimulation’ (Rattay 1990). He is also a pioneer in simulations for retina implants (Rattay et al. 2003) and spinal cord stimulation for locomotion (Rattay et al. 2000). Besides many other discoveries, Rattay showed that chronaxie, the classical measure for neuronal excitability, is not constant (~ 0.7 x the time constant of the membrane was assumed since 1932), but it becomes about 10 times shorter for extra- vs. intracellular stimulation if the electrode is close to the neuron (Rattay et al. 2012). In 2013, a biography of Rattay with interesting scientific details on electrical stimulation fundamentals appeared in the pioneer series of the journal Artificial Organs. https://doi.org/10.1111/aor.12258 

Frank Rattay supervised 156 Master and 76 PhD theses, most of them in the field of Biomedical Engineering. In 2008 he initiated the Master studies ‘Biomedical Engineering’ at TU Wien.

Hochmair-Desoyer, I. J., Hochmair, E. S., Motz, H., Rattay, F. 1984. A model for the electrostimulation of the nervus acusticus. Neuroscience, 13(2), 553-562. https://doi.org/10.1016/0306-4522(84)90248-3

Motz, H., Rattay, F. 1986. A study of the application of the Hodgkin-Huxley and the Frankenhaeuser-Huxley model for electrostimulation of the acoustic nerve. Neuroscience, 18(3), 699-712. doi: 10.1016/0306-4522(86)90064-3

Rattay, F., 1986. Analysis of models for external stimulation of axons. IEEE Trans Biomed Eng 33, 974–7. doi: 10.1109/TBME.1986.325670

Rattay, F., 1990. Electrical Nerve Stimulation: Theory, Experiments and Applications. Springer, New York. 

Rattay, F., Minassian, K., Dimitrijevic, M. R. 2000. Epidural electrical stimulation of posterior structures of the human lumbosacral cord: 2. quantitative analysis by computer modeling. Spinal cord, 38(8), 473-489. https://doi.org/10.1038/sj.sc.3101039

Rattay, F., Greenberg, R.J., Resatz, S., 2003. Handbook of Neuroprosthetic Methods. CRC Press, Boca Raton FL, pp. 39–71.

Rattay, F., Paredes, L.P., Leao, R.N., 2012. Strength-duration relationship for intra- versus extracellular stimulation with microelectrodes. Neuroscience 214, 1–13. https://doi.org/10.1016/j.neuroscience.2012.04.004

Reference

Fellner A, Wenger C, Heshmat A, Rattay F. Auditory nerve fiber excitability for alternative electrode placement in the obstructed human cochlea: electrode insertion in scala vestibuli versus scala tympani. J Neural Eng. 2024;21(4). doi: 10.1088/1741-2552/ad6597. 

Go to J Neural Eng.

Check Also

Experimental Observation of Electromagnetically Induced Transparency-Like Effect in a Hybrid Plasmonic Multilayer System