Maximising Effective Audibility in Hearing Aid Fitting Teresa Ching, Harvey
Dillon, Richard Katsch and Denis Byrne Hearing loss causes some parts of the speech spectrum to be inaudible, making it difficult for hearing impaired people to understand speech. If speech is amplified so that the short term rms levels in every frequency band are 30 dB or more above threshold, audibility is perfect, and according to the Speech Intelligibility Index (SII) model, speech intelligibility is maximised. This is not always the case for people with severe or profound hearing impairment. For them, hearing more speech can mean understanding less, because the ear cannot make effective use of all the information that exceeds their thresholds at each frequency. This is especially more so at the high frequencies than at the low frequencies. We have derived a relationship between effective audibility and physical audibility at different frequencies from experimental data. This was used for modifying the SII, so that the model could be adopted to determine the optimal gain at different frequencies for maximising intelligibility for people with different degrees of hearing loss. This paper elucidates how the principle has been applied in deriving a hearing aid prescription procedure, illustrates how a prescription that aims to maximise effective audibility differs from one that maximises audibility, and demonstrates that the former is better for speech intelligibility than the latter. |