Baer Thomas, Moore Brian C J, Kluk Karolina
Department of Experimental Psychology, University of Cambridge, United Kingdom.
J Acoust Soc Am. 2002 Sep;112(3 Pt 1):1133-44. doi: 10.1121/1.1498853.
People with high-frequency sensorineural hearing loss differ in the benefit they gain from amplification of high frequencies when listening to speech. Using vowel-consonant-vowel (VCV) stimuli in quiet that were amplified and then low pass filtered with various cutoff frequencies, Vickers et aL [J. Acoust. Soc. Am. 110, 1164-1175 (2001)] found that the benefit from amplification of high-frequency components was related to the presence or absence of a cochlear dead region at high frequencies. For hearing-impaired subjects without dead regions, performance improved with increasing cutoff frequency up to 7.5 kHz (the highest value tested). Subjects with high-frequency dead regions showed no improvement when the cutoff frequency was increased above about 1.7 times the edge frequency of the dead region. The present study was similar to that of Vickers et al. but used VCV stimuli presented in background noise. Ten subjects with high-frequency hearing loss, including eight from the study of Vickers et al., were tested. Five had dead regions starting below 2 kHz, and five had no dead regions. Speech stimuli at a nominal level of 65 dB were mixed with spectrally matched noise, amplified according to the "Cambridge" prescriptive formula for each subject and then low pass filtered. The noise level was chosen separately for each subject to give a moderate reduction in intelligibility relative to listening in quiet. For subjects without dead regions, performance generally improved with increasing cutoff frequency up to 7.5 kHz, on average more so in noise than in quiet. For most subjects with dead regions, performance improved with cutoff frequency up to 1.5-2 times the edge frequency of the dead region, but hardly changed with further increases. Calculations of speech audibility using a modified version of the articulation index showed that application of the Cambridge formula was at least partially successful in making high-frequency components of the speech audible for subjects with dead regions, and that such subjects often failed to benefit from increased audibility of the speech at high frequencies.
高频感音神经性听力损失患者在通过高频放大来聆听语音时所获得的益处存在差异。使用在安静环境中经放大后再以不同截止频率进行低通滤波的元音 - 辅音 - 元音（VCV）刺激，维克斯等人[《美国声学学会杂志》110, 1164 - 1175 (2001)]发现，高频成分放大带来的益处与高频处是否存在耳蜗死区有关。对于没有死区的听力受损受试者，截止频率增加到7.5千赫（测试的最高值）时，表现会有所改善。高频有死区的受试者，当截止频率增加到高于死区边缘频率的约1.7倍时，表现没有改善。本研究与维克斯等人的研究类似，但使用了在背景噪声中呈现的VCV刺激。测试了10名高频听力损失患者，其中8名来自维克斯等人的研究。5名患者的死区起始于2千赫以下，5名患者没有死区。标称水平为65分贝的语音刺激与频谱匹配的噪声混合，根据每个受试者的“剑桥”规定公式进行放大，然后进行低通滤波。为每个受试者分别选择噪声水平，以使可懂度相对于安静聆听时有适度降低。对于没有死区的受试者，截止频率增加到7.5千赫时，表现通常会改善，平均而言在噪声环境中比在安静环境中改善得更多。对于大多数有死区的受试者，截止频率增加到死区边缘频率的1.5 - 2倍时，表现会改善，但进一步增加时变化不大。使用改进版清晰度指数进行的语音可听度计算表明，应用剑桥公式至少在一定程度上成功地使有死区的受试者能够听到语音的高频成分，并且这类受试者往往无法从高频语音可听度的提高中受益。
