Rader Tobias, Döge Julia, Adel Youssef, Weissgerber Tobias, Baumann Uwe
Audiological Acoustics, ENT Department, University Hospital Frankfurt, Germany.
Audiological Acoustics, ENT Department, University Hospital Frankfurt, Germany.
Hear Res. 2016 Sep;339:94-103. doi: 10.1016/j.heares.2016.06.013. Epub 2016 Jul 1.
In normal hearing, the pitch of an acoustic tone can theoretically be encoded by either the place of stimulation in the cochlea or the corresponding rate of vibration. Thus spectral attributes and temporal fine structure of an acoustic signal are naturally correlated. Cochlear implants (CIs), neural prosthetic devices that restore hearing in the profoundly hearing impaired, currently disregard this mechanism; electrical stimulation is provided at fixed electrode positions with default place independent stimulation rate assignments. This does not account for individual cochlear encoding depending on electrode array placement, variations in insertion depth, and the proximity to nerve fibers. Encoding pitch in such manner delivers limited tonal information. Consequently, music appraisal in CI users is often rated cacophonic while speech perception in quiet is close to normal in top performers. We hypothesize that this limitation in electric stimulation is at least partially due to the mismatch between frequency and place encoding in CIs. In the present study, we determined individual electrode locations by analysis of cochlear radiographic images obtained after surgery and calculated place dependent stimulation rates according to models of the normal tonotopic function. Pitch matching in CI users with single-sided deafness shows that place dependent stimulation rates allow thus far unparalleled restoration of tonotopic pitch perception. Collapsed data of matched pitch frequencies as a function of calculated electrical stimulation rate were well fitted by linear regression (R(2) = 0.878). Sound processing strategies incorporating place dependent stimulation rates are expected to improve pitch perception in CI users.
在正常听力情况下，理论上声音的音高可以通过耳蜗中的刺激位置或相应的振动频率来编码。因此，声音信号的频谱属性和时间精细结构自然是相关的。人工耳蜗（CI）是一种为极重度听力受损者恢复听力的神经假体装置，目前忽略了这一机制；电刺激是在固定电极位置以默认的与位置无关的刺激速率分配来提供的。这没有考虑到根据电极阵列放置、插入深度变化以及与神经纤维的接近程度而产生的个体耳蜗编码情况。以这种方式编码音高所传递的音调信息有限。因此，人工耳蜗使用者对音乐的评价往往被认为是刺耳的，而在表现最佳的使用者中，安静环境下的言语感知接近正常。我们推测，这种电刺激的局限性至少部分是由于人工耳蜗中频率和位置编码之间的不匹配所致。在本研究中，我们通过分析术后获得的耳蜗X线图像来确定个体电极位置，并根据正常音调定位功能模型计算与位置相关的刺激速率。单侧耳聋的人工耳蜗使用者的音高匹配结果表明，与位置相关的刺激速率能够实现迄今为止无与伦比的音调定位音高感知恢复。匹配音高频率的汇总数据作为计算出的电刺激速率的函数，通过线性回归得到了很好的拟合（R² = 0.878）。预计采用与位置相关的刺激速率的声音处理策略将改善人工耳蜗使用者的音高感知。
