High-frequency audibility: Benefits for hearing-impaired listeners
High-frequency audibility: Benefits for hearing-impaired listeners
复制标题
DOI：
10.1121/1.423247
复制
发表时间：
1998-07-01
影响因子：
2.4
通讯作者：
Turner, CW
中科院分区：
文献类型：
--
作者：
Hogan, CA;Turner, CW
本研究是一个系统的调查提供听觉受损的听众可听到的高频语音信息的好处。五名听力正常的听众和九名高频听力受损的听众识别出了在一些截止频率下经过低通过滤的无意义音节。作为量化每种情况下可听性的一种手段，我们计算了每个听者在每种情况下的发音指数（Articulation Index， AI）。大多数听力受损的听众在提供额外的可听高频信息后表现出语音识别的改善。在某些情况下，对于受损更严重的听者，增加高频语音信息的可听性不会导致语音识别的进一步改善，甚至会导致语音识别的下降。一种新的衡量听力受损听众在特定频段内使用信息的能力的方法被称为“效率”。这种方法比较了为听力受损的听者提供一定程度的言语可听性提高所带来的好处，以及在听力正常的听者中观察到的同样程度的言语可听性提高所带来的好处。使用旧的人工智能方法和新的人工智能方法（考虑到高语音呈现水平的影响）计算效率。结果中有一个清晰的模式表明，当给定频率的听力损失程度超过55 dB HL时，为该频率区域提供额外可听性的效果就会减弱，特别是当这种听力损失程度出现在4000 Hz及以上的频率时。“旧”和“新”人工智能程序的分析比较表明，这些听众的一些（但不是全部）语音识别缺陷是由于高表达水平造成的。(C) 1998美国声学学会。
The present study was a systematic investigation of the benefit of providing hearing-impaired listeners with audible high-frequency speech information. Five normal-hearing and nine high-frequency hearing-impaired listeners identified nonsense syllables that were low-pass filtered at a number of cutoff frequencies. As a means of quantifying audibility for each condition, Articulation Index (AI) was calculated for each condition for each listener. Most hearing-impaired listeners demonstrated an improvement in speech recognition as additional audible high-frequency information was provided. In some cases for more severely impaired listeners, increasing the audibility of high-frequency speech information resulted in no further improvement in speech recognition, or even decreases in speech recognition. A new measure of how well hearing-impaired listeners used information within specific frequency bands called "efficiency" was devised. This measure compared the benefit of providing a given increase in speech audibility to a hearing-impaired listener to the benefit observed in normal-hearing listeners for the same increase in speech audibility. Efficiencies were calculated using the old AI method and the new Al method (which takes into account the effects of high speech presentation levels). There was a clear pattern in the results suggesting that as the degree of hearing loss at a given frequency increased beyond 55 dB HL, the efficacy of providing additional audibility to that frequency region was diminished, especially when this degree of hearing loss was present at frequencies of 4000 Hz and above. A comparison of analyses from the "old" and "new" AI procedures suggests that some, but not all, of the deficiencies of speech recognition in these listeners was due to high presentation levels. (C) 1998 Acoustical Society of America.