当前位置: X-MOL 学术Atten. Percept. Psychophys. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Dichotic listening performance with cochlear-implant simulations of ear asymmetry is consistent with difficulty ignoring clearer speech
Attention, Perception, & Psychophysics ( IF 1.7 ) Pub Date : 2021-03-29 , DOI: 10.3758/s13414-021-02244-x
Matthew J. Goupell , Daniel Eisenberg , Kristina DeRoy Milvae

There are an increasing number of bilateral and single-sided-deafness cochlear-implant (CI) users who hope to achieve improved spatial-hearing abilities through access to sound in both ears. It is, however, unclear how speech is processed when inputs are functionally asymmetrical, which may have an impact on spatial-hearing abilities. Therefore, functionally asymmetrical hearing was controlled and parametrically manipulated using a channel vocoder as a CI simulation. In Experiment 1, normal-hearing (NH) listeners performed a dichotic listening task (i.e., selective attention to one ear, ignoring the other) using asymmetrical signal degradation. Spectral resolution varied independently in each ear (4, 8, 16 channels, and unprocessed control). Performance decreased with decreasing resolution in the target ear and increasing resolution in the interferer ear. In Experiment 2, these results were replicated using a divided attention task (attend to both ears, report one after sentence completion) in both NH and bilateral CI listeners, although overall performance was lower than in Experiment 1. In Experiment 3, frequency-to-place mismatch simulated shallow CI insertion depths (0, 3, 6-mm shifts, and unprocessed control). Performance mostly decreased with increasing shift in the target ear and decreasing shift in the interferer ear; however, performance nonmonotonicities occurred. The worst performance occurred when the shift matched across ears, suggesting that pitch similarity increases difficulty. The results show that it is more difficult to attend an ear that is relatively degraded or distorted, which may set spatial-hearing limitations for CI users when trying to attend to a target in complex auditory scenes.



中文翻译:

耳不对称的人工耳蜗模拟的双耳听觉表现与忽略清晰语音的困难一致

越来越多的双边和单侧耳聋人工耳蜗(CI)用户希望通过获得双耳的声音来提高空间听觉能力。但是,尚不清楚当输入在功能上不对称时如何处理语音,这可能会影响空间听觉能力。因此,使用通道声码器作为CI模拟来控制和参数化功能上不对称的听力。在实验1中,正常听力(NH)的听众使用不对称信号衰减来执行双耳听觉任务(即选择性注意一只耳朵,而忽略另一只耳朵)。每只耳朵的光谱分辨率均独立变化(4、8、16个通道和未处理的对照)。随着目标耳分辨率的降低和干扰耳分辨率的提高,性能会下降。在实验2中,虽然总体表现低于实验1,但在NH和双边CI听众中,使用分散注意力的任务(听着双耳,在句子完成后报告一个)重复了这些结果,尽管总体表现低于实验1。位失配模拟浅CI插入深度(偏移0、3、6毫米,以及未处理的控制)。随着目标耳偏移的增加和干扰耳偏移的减少,性能大多下降。但是,出现了性能非单调性。当移位跨耳朵匹配时,性能最差,这表明音高相似度增加了难度。

更新日期:2021-03-30
down
wechat
bug