Impact of Mismatched Spectral Amplitude Levels on Vowel Identification in Simulated Electric-acoustic Hearing

Abstract

The benefits of combined electric-acoustic stimulation (EAS) in terms of better speech recognition have been well documented in the literature for patients fitted with a hearing aid and a cochlear implant, providing them low-frequency and high-frequency speech information, respectively. This work assessed the effect of mismatched spectral amplitude levels on vowel identification in simulated EAS hearing. The spectral amplitude levels of four synthetic vowels (i.e., /iy/, /eh/, /oo/ and /ah/) were modified to amplify either low-frequency (≤600 Hz) or high-frequency (>600 Hz) portion, and the EAS-processed stimuli were presented to normal-hearing listeners to identify. Results showed declined vowel identification scores in response to acoustic or electric spectral amplitude amplification, and the specific loudness pattern computed from Moore et al.’s model was found to effectively account for the variance of vowel identification scores.