Abstract：
Cochlear implants are medical devices that have restored hearing to approximately one million people around the world. Outcomes are impressive and most recipients attain excellent speech comprehension in quiet without relying on lip-reading cues, but pitch resolution is poor compared to normal hearing. Amplitude modulation of electrical stimulation is a primary cue for pitch perception in cochlear implant users. The experiments described in this article focus on the relationship between sensitivity to amplitude modulations and pitch resolution based on changes in the frequency of amplitude modulations. In the first experiment, modulation sensitivity and pitch resolution were measured in adults with no known hearing loss and in cochlear implant users with sounds presented to and processed by their clinical devices. Stimuli were amplitude-modulated sinusoids and amplitude-modulated narrow-band noises. Modulation detection and modulation frequency discrimination were measured for modulation frequencies centered on 110, 220, and 440 Hz. Pitch resolution based on changes in modulation frequency was measured for modulation depths of 25 %, 50 %, 100 %, and for a half-waved rectified modulator. Results revealed a strong linear relationship between modulation sensitivity and pitch resolution for cochlear implant users and peers with no known hearing loss. In the second experiment, cochlear implant users took part in analogous procedures of modulation sensitivity and pitch resolution but bypassing clinical sound processing using single-electrode stimulation. Results indicated that modulation sensitivity and pitch resolution was better conveyed by single-electrode stimulation than by clinical processors. Results at 440 Hz were worse, but also not well conveyed by clinical sound processing, so it remains unclear whether the 300 Hz perceptual limit described in the literature is a technological or biological limitation. These results highlight modulation depth and sensitivity as critical factors for pitch resolution in cochlear implant users and characterize the relationship that should inform the design of modulation enhancement algorithms for cochlear implants.
摘要：
人工耳蜗是一种医疗设备，可以恢复全世界大约一百万人的听力。结果令人印象深刻，大多数接受者在安静的情况下获得了出色的言语理解，而无需依靠唇读提示，但是与正常听力相比，音高分辨率较差。电刺激的幅度调制是耳蜗植入物用户的音高感知的主要线索。本文描述的实验集中在对幅度调制的灵敏度与基于幅度调制频率变化的音调分辨率之间的关系上。在第一个实验中,在没有已知听力损失的成年人和人工耳蜗使用者中测量了调制灵敏度和音调分辨率，并将声音呈现给他们的临床设备并通过其处理。刺激是幅度调制的正弦曲线和幅度调制的窄带噪声。对于以110、220和440Hz为中心的调制频率，测量了调制检测和调制频率鉴别。根据调制频率的变化测量了25％的调制深度的音调分辨率，50%,100%,和半波整流调制器。结果表明，对于没有已知听力损失的人工耳蜗使用者和同伴，调制灵敏度与音调分辨率之间存在很强的线性关系。在第二个实验中,人工耳蜗使用者参与了调制灵敏度和音调分辨率的类似程序，但绕过了使用单电极刺激的临床声音处理。结果表明，单电极刺激比临床处理器更好地传达了调制灵敏度和音调分辨率。440Hz时的结果更差，但临床声音处理也不能很好地传达，因此，目前尚不清楚文献中描述的300Hz感知限制是技术限制还是生物学限制。这些结果突出了调制深度和灵敏度是耳蜗植入物用户的间距分辨率的关键因素，并表征了应该为耳蜗植入物的调制增强算法设计提供信息的关系。
