Intelligibility of vowels produced by post-lingually deafened cochlear implant users

The present study addresses the effect of cochlear implantation on the intelligibility of vowels produced by 20 post-lingually deafened Dutch subjects. All subjects received the Nucleus-22 cochlear implant (3 WSP and 17 MSP processors). Speech recordings were made pre-implantation and three and twelve months post-implantation with the implant switched on and off. Vowel intelligibility (monophthongs only) was determined using a panel of listeners. For all implanted subjects intelligibility was measured in a noisy background. For seven poorly speaking subjects it was also measured in a quiet background. After implantation with the Nucleus-22 device the results showed that vowel intelligibility, measured for all subjects in a noisy background, increased for most of them (about 15), while it increased for about half the number of poorly speaking subjects measured in a quiet background. Twelve months after implantation vowel intelligibility, measured for all subjects in noise, appeared to be based on first and second formant information. This was also found for the subgroup of seven subjects performing poorly pre-implantation when analysed separately. However, vowel intelligibility for this subgroup, when measured in a quiet background, was based also on vowel duration. The differences between the overall result in noise and the results of the subgroup in quiet should be attributed mainly to the noise and not to aspects of poor speech production in the subgroup. In addition, this study addresses the relationship between the intelligibility scores and objective measurements of vowel quality performed in a previous study. The results showed that the vowel intelligibility scores are mainly determined by the position of the second formant frequencies.