Skip to main content

How do low frequency acoustic cues improve speech recognition and music appreciation for cochlear implant users?

Final Report Summary - LFAA (How do low frequency acoustic cues improve speech recognition and music appreciation for cochlear implant users?)

Key Results

• Identified cues which improve cochlear implant speech-in-noise understanding
• Advanced our understanding of neural processing of complex sounds by cochlear implant users
• Developed cutting edge technologies to easily measure and assess brain response in cochlear implant users

Impact

• Better speech understanding in noise for cochlear implant users
• Optimize cochlear implant fitting process, improving outcomes and reducing work load
• Support and development for CI research in Ireland.
• Commerilization of technologies, leading to economic growth


Cochlear implants (CIs) are the most successful neural prosthesis that has yet been developed. Routine cochlear implantation began in the 1970s and since then CIs have helped restore hearing in over 220,000 deaf people world-wide. Advances in speech processing strategies and implant design mean that most CI users, if implanted as prelingual children or as postlingually deafened adults, can understand speech in a quiet environment and can converse over the telephone. However, to achieve this level of speech perception the CI requires extensive initial fitting and frequent tuning. As the number of implant users increase, so does the work load for medical professionals. Another significant problem for CI users is speech recognition in noise and understanding multi-talker speech. Recent studies have shown that CI users with residual low frequency hearing have both better speech recognition in noise and improved melody recognition. The aim of this project was to use neural, EEG methods, to understand these issues and improve cochlear implant outcomes.

The project’s first study examined the cues that underlie speech-in-noise perception in CI users and revealed that, in addition to a low frequency acoustic cue, CI uses could make use of differences in stimulation rate to improve speech-in-noise perception. A speech perception model showed that this improvement was probably due to a listening in the gaps cue rather that a temporal pitch cue. This new knowledge will facilitate the design of cochlear implant speech processing strategies which perform better in noisy environments

To study the relationship between behavioral outcomes, such as speech perception, and the underlying neural mechanism a paired EEG – psychoacoustic testing method was developed and evaluated in CI and normal hearing subjects. A good correlation with behavioral and neural thresholds was found for a number of different stimuli. These results shed light on the neural processing of these complex stimuli and indicate a cortical origin for the processing.

While measuring EEG in CI subjects two new methods of significant clinical relevance were developed. One allows a standard CI to be used as a recording device to measure EEG, the other is a new method to separate the EEG signal from the CI related artifact. It has been shown that combining the CI recording technique with the spectrally rippled stimuli provides a clinically significant evaluation procedure: by simply pressing a button on the CI an audiologist can get a quick estimate of a user’s speech perception. This will greatly improve the treatment and follow offered to CI users.

This new technology means that it may now be possible to develop a closed-loop cochlear implant (see inset). Instead of the audiologist fitting the implant in an open-loop way, as is currently done, a closed-loop CI could automatically measure neural response and optimize stimulation accordingly.