Previous studies have indicated that individuals with cochlear hearing loss perform poorly in concurrent vowel identification task. This indicates that individuals with cochlear hearing loss do not use F0 cues to segregate two acoustic streams, as much as normal hearing individuals do. However, which of vowel features (place, tongue height etc.,) are better transmitted when the F0s of two concurrently presented vowels are varied is not known. Moreover, the contribution of stream segregation abilities in understanding of speech in the presence of competing signal is also not clear. Hence, this study aimed to evaluate the relationship between concurrent vowel identification scores and speech perception abilities in noise in individuals with cochlear hearing loss and compared that to normal hearing individuals. Specifically, we measured the identification of vowels /e/, /i/, /o/ and /u/ when vowel /a/ was presented simultaneously as a competing stimulus in 14 individuals with normal hearing and 15 individuals with cochlear hearing loss. Vowel identification scores were measured in four conditions: with 0 semitone difference between fundamental frequencies of concurrent vowels, with 1 semitone, 2 semitone and 4 semitone difference in the fundamental frequencies of concurrent vowels. Signal to noise ratio required to identify 50% of presented speech (SNR-50) in presence of four talker babble was also measured using standardized sentence list. Furthermore, relationship between concurrent vowel identification and SNR-50 was also evaluated. The results of the present study showed that individuals with cochlear hearing loss had poorer concurrent vowel identification scores compared to the normal hearing group, especially in 0 semitone difference condition. There were significant correlations between concurrent vowel identification scores and SNR-50 indicating, differences in the F0 are one of the most robust cues for stream segregation.
All Science Journal Classification (ASJC) codes
- Acoustics and Ultrasonics