Related Experiment Video
Updated: Oct 13, 2025
03:58
Enhancing Electrode Location Assessment in Cochlear Implantation via Computed Tomography Image Fusion
Published on: January 17, 2025
586
Valid Acoustic Models of Cochlear Implants: One Size Does Not Fit All
Mario A Svirsky1,2,3, Nicole Hope Capach1,2, Jonathan D Neukam1,2
1New York University.
Summary
Valid acoustic models for cochlear implants (CI) were found using personalized frequency mismatch. These models closely mimic CI sound perception and speech understanding, validating their use in research.
Area of Science:
- Auditory Neuroscience
- Biomedical Engineering
- Speech Processing
Background:
- Noise or tone vocoders are widely used as acoustic models for cochlear implants (CI).
- However, these vocoder models have lacked convincing validation in previous studies.
Purpose of the Study:
- To test if tone or noise vocoders can accurately model cochlear implant (CI) sound perception and speech understanding.
- To validate the use of vocoders as acoustic models for CIs.
- To investigate if personalized frequency mismatch is necessary for accurate acoustic CI modeling.
Main Methods:
- Acoustic models were evaluated by single-sided deaf CI users.
- Participants compared sound from their CI in one ear to an acoustic model in the other.
- Evaluated models included frequency-matched and self-selected models with individualized frequency mismatch.
Main Results:
- Self-selected acoustic models demonstrated speech perception levels comparable to the CI.
- These personalized models matched the CI in perceptual quality, including intelligibility, harshness, and pleasantness.
- Individualized frequency mismatch improved model validity.
Conclusions:
- Validated acoustic models for cochlear implants (CI) exist and differ from commonly used ones.
- Personalized frequency mismatch is crucial for optimizing the accuracy of CI acoustic models.
- This finding may relate to the basalward frequency shift observed in post-lingual CI users.
