Phoneme recognition by deaf individuals using the multichannel nucleus cochlear implant

Area of Science:

• Audiology
• Speech Science
• Biomedical Engineering

Background:

• Cochlear implants are vital for restoring hearing in deaf individuals.
• Understanding phoneme identification cues is crucial for improving cochlear implant technology.
• The Nucleus 22-channel cochlear implant's speech processing strategy influences auditory perception.

Purpose of the Study:

• To determine the acoustic cues used by deaf cochlear implant users for phoneme identification.
• To analyze the role of voicing and formant frequencies in speech perception.
• To evaluate the impact of electrode array insertion depth on vowel recognition.

Main Methods:

• Open-set speech audiometry was conducted on five deaf Nucleus 22-channel cochlear implant users.
• Phoneme confusions were analyzed using confusion matrices and Kruskal analysis.
• Speech material included Dutch Consonant-Vowel-Consonant (CVC) words.

Main Results:

• Consonant recognition primarily relies on the voicing feature.
• Vowel identification is based on the first and second formant frequencies.
• Partial electrode insertion limited second formant information, impacting vowel recognition.

Conclusions:

• The Nucleus speech processor's coding strategy aligns with observed phoneme identification cues.
• Kruskal analysis of phoneme confusions can inform cochlear implant programming and performance evaluation.
• Voicing and formant frequency processing are key to speech understanding with cochlear implants.