Sound localization in noise by normal-hearing listeners and cochlear implant users.
Sound localization in noise by normal-hearing listeners and cochlear implant users.
复制标题
DOI：
10.1097/aud.0b013e318257607b
复制
发表时间：
2012-07
期刊：
影响因子：
3.7
通讯作者：
Seeber BU
中科院分区：
文献类型：
--
作者：
Kerber S;Seeber BU
本研究旨在表征不同信噪比（SNR）下干扰噪声中的水平平面声音定位，并比较听力正常的听众和单侧和双侧人工耳蜗植入体（CI）用户的性能。CI用户报告在嘈杂的环境中收听困难。虽然他们在言语理解方面的困难已经在几项研究中进行了调查，但在背景噪声中定位声音的能力尚未得到广泛的研究，尽管双耳听力在嘈杂环境中的益处最大。声音定位是双耳处理的一种测量方法，因此非常适合评估双侧植入的益处。这些结果将告知临床医生和植入物制造商如何集中精力在嘈杂环境中改善CI的定位。六个正常听力的听众，四个单边和十个双边CI用户表示使用光指针方法感知声源的位置。目标声音是从11个扬声器中的一个播放的噪声脉冲，这些扬声器放置在自由场中的正面水平面中的−80°和+80°之间。定位在安静和弥漫性背景噪声中进行评估，SNR在+10和-7 dB之间。语音接收阈值进行了测量，并检查其与本地化结果的关系。定位性能随着SNR的降低而下降：目标声音被感知到更接近正中平面，并且响应的标准差增加。使用“空间分辨率”（SR）的测量来比较各组之间的定位性能。该测量给出了两个声源之间的角距，使理想的观察者能够在69.1%的时间内正确区分它们。对于所有的参与者SR增加降低SNR，即在低SNR的声源之间的空间分离保持可区分的，只有当它是较大的。听力正常的参与者表现最好，安静时SR在1.4°到5.1°之间。双侧CI用户在安静时显示SR在8.3°和43.6°之间，大约相当于SNR为−5 dB时正常听力听者的空间分辨率。大多数双侧CI用户在SNR为-3 dB时无法正确确定声音来自哪一侧。总体而言，SNR必须至少为+7 dB，以实现接近所有双边CI用户在安静时的定位性能。空间分辨率和语音接收阈值之间没有显着的相关性，但语音处理器的灵敏度设置显着影响性能。单侧CI用户表现出最严重的定位问题，只有四分之二的参与者能够正确地确定哪一边的声音来自安静。这项研究是第一次检查声音定位与CI在不同的SNR，并将其与正常听力。结果证实，本地化与CI强烈干扰嘈杂的情况下。在安静和嘈杂的情况下，双侧CI明显上级单侧CI。对于双侧CI，定位在中等高的绝对噪声水平（> 63 dB SPL）下下降，这表明将声学动态范围扩展到更高的水平将是有益的。语言接收阈值和空间分辨率之间缺乏相关性，这突出表明需要进行额外的临床试验来评估第二个植入物的双耳受益。
This study aimed to characterize horizontal-plane sound localization in interfering noise at different signal-to-noise ratios (SNRs) and to compare performance across normal-hearing listeners and users of unilateral and bilateral cochlear implants (CIs). CI users report difficulties with listening in noisy environments. While their difficulties with speech understanding have been investigated in several studies, the ability to localize sounds in background noise has not extensively been examined, despite the benefits of binaural hearing being greatest in noisy situations. Sound localization is a measure of binaural processing and is thus well suited to assessing the benefit of bilateral implantation. The results will inform clinicians and implant manufacturers how to focus their efforts to improve localization with CIs in noisy situations. Six normal-hearing listeners, four unilateral and ten bilateral CI users indicated the perceived location of sound sources using a light pointer method. Target sounds were noise pulses played from one of 11 loudspeakers placed between −80° and +80° in the frontal horizontal plane in the free-field. Localization was assessed in quiet and in diffuse background noise at SNRs between +10 and −7 dB. Speech reception thresholds were measured and their relation to the localization results examined. Localization performance declined with decreasing SNR: Target sounds were perceived closer to the median plane and the standard deviation of responses increased. Localization performance across groups was compared using a measure of “Spatial Resolvability” (SR). This measure gives the angular separation between two sound sources that would enable an ideal observer to correctly distinguish them 69.1% of the time. For all participants SR increased with decreasing SNR, i.e. at low SNRs the spatial separation between sound sources remained distinguishable only when it was larger. Normal-hearing participants performed best, with SR between 1.4° and 5.1° in quiet. Bilateral CI users showed SR between 8.3° and 43.6° in quiet, corresponding approximately to the spatial resolution of normal-hearing listeners at an SNR of −5 dB. Most bilateral CI users had lost the ability to correctly determine which side the sound came from at an SNR of −3 dB. Overall, the SNR had to be at least +7 dB to achieve localization performance near to that in quiet for all bilateral CI users. No significant correlation was found between spatial resolution and speech reception thresholds, but the speech-processor sensitivity setting did significantly affect performance. Unilateral CI users showed the most severe localization problems, with only two of four participants being able to correctly determine which side sounds came from in quiet. This study is the first to examine sound localization with CIs at various SNRs and to compare it to normal hearing. The results confirm that localization with CIs is strongly disrupted in noisy situations. Bilateral CIs were shown to be clearly superior over unilateral CIs for localization in quiet and in noisy situations. With bilateral CIs localization declined at moderately high absolute noise levels (> 63 dB SPL), suggesting that an extension of the acoustic dynamic range to higher levels would be beneficial. The absence of a relation between speech reception thresholds and spatial resolution highlights the need for additional clinical tests to assess the binaural benefit of a second implant.