Fine-Structure-Based Models for Cochlear Implant Advancement

• Applicants: Dr. Siwei Bai, Ph.D.; Professor Dr.-Ing. Werner Hemmert
• Subject Area: Otolaryngology, Phoniatrics and Audiology ; Acoustics ; Cognitive, Systems and Behavioural Neurobiology ; Medical Physics, Biomedical Technology
• Term: since 2018
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 415658392

Cochlear implants (CIs) are the most successful neuroprostheses: they restore hearing in deaf people to a surprisingly high degree. Nevertheless, there is still a large variability in outcomes, especially a lack in understanding the deficits of poor performers. As direct recordings in humans are not possible, computational models that predict neuronal excitation patterns are fundamental to advance this technology. We have therefore developed anatomically accurate cochlea models from high-resolution micro-CT scans of various cochleas from deceased human subjects (total of 8) during the last funding period and established a processing pipeline to derive excitation patterns of the auditory nerve. During this funding period, we will investigate the effects of degeneration of peripheral axons including the distribution of functionally important voltage-gated ion channels with the following work packages: 1) Morphometric measurements of type I spiral ganglion neurons (SGNs) in different tonotopic regions in cochleas from normal hearing- as well as hearing impaired donors. The distribution of voltage-gated ion channels will be obtained from our post-mortem human specimens based on immunostainings and serial block-face electron microscopy to characterise innervation patterns and the exact localization of different SGN subtypes. 2) Investigation of sensorineural degeneration patterns based on our specimens and correlation with audiometric data. These patterns will be replicated in the computational model, and their effects on the quality of speech coding will be studied. 3) Assembly of all findings in our collection of human cochlea models, incorporating a realistic distribution of relevant ion channels. We will also, for the first time, include dominant temporal effects like adaptation processes found in electrical stimulation, which are currently missing in existing biophysical models of SGNs. These processes are also important for all other interventions that rely on electrical nerve stimulation. In the long run, this project will provide the basis for personalised, user-specific stimulation and coding strategies. Products based on our interdisciplinary research, such as a new vocoder that matches the hearing outcome of CI users, will help us acquire better knowledge of the performance, which will ultimately lead to the advancement of CI technology. With our models we will vice versa be able to investigate the possibility to diagnose degeneration patterns in CI users based on the combination of quasi-monopolar or tripolar pulses and multipolar electrode configurations, which will be a big step towards customised coding strategies. In addition, our evaluation of typical degeneration patterns of SGNs and sensory cells, especially inner hair cells, and their correlation to audiograms will provide more information on the mechanisms of hearing loss, which unfortunately becomes a more and more severe burden in our ageing society.

• DFG Programme: Research Grants
• International Connection: Austria
• Partner Organisation: Fonds zur Förderung der wissenschaftlichen Forschung (FWF)
• Cooperation Partner: Privatdozent Dr. Rudolf Glückert