新人工耳蜗使用者采用耳蜗拓扑拟合时精细结构编码的评估：一项前瞻性、随机、双盲、交叉研究

文献信息

中文摘要

多项研究已证明耳蜗拓扑拟合对新人工耳蜗（CI）使用者噪声中言语感知的益处。然而，这些研究使用了不同的编码策略（CIS、HDCIS、FSP、FS4或混合）和不同的频谱。本研究旨在评估精细结构（FS）线索对采用耳蜗拓扑拟合图的新CI使用者在噪声和安静环境中言语感知以及音乐感知的影响。

一项前瞻性、随机、双盲、2周期交叉研究于2023年4月至2024年6月进行。所有参与者均通过平板计算机断层扫描（CT）和基于Greenwood函数的重建软件确定耳蜗拓扑拟合。在单一三级转诊中心招募了24名新CI使用者，年龄18岁或以上，患有双侧重度至极重度感音神经性听力损失或完全听力损失少于5年。参与者被随机分配至先FS后常规编码，或相反顺序，为期6周，每期后进行言语和音乐评估。

FS线索显著改善了高信噪比下的噪声中言语识别（SNR = +9 dB，平均效应（ME）= 6.0%，标准化效应量（SES）= 0.62，p = 0.010）。在SNR超过+6 dB时观察到周期效应（SNR = +9 dB，p = 0.010；SNR = +6 dB，p = 0.018），但无统计学显著的组效应。安静环境中的言语识别和纯音测听未受FS编码影响。FS线索增强了旋律轮廓识别（ME = 7.2%，SES = 0.51，p = 0.030），并且在探索性亚组分析中，低频音高辨别也得到改善（ME = 8.5 Hz，SES = 0.72，p = 0.04）。超过75%的参与者在研究完成后使用FS编码保留了耳蜗拓扑图。

在耳蜗拓扑图上使用FS编码策略可能在有利SNR下改善噪声中言语识别，而不损害安静环境中的言语理解。它还可能增强选定的复杂声音感知方面，特别是音高和轮廓相关处理，而对主观音乐感知的益处似乎不太一致。

英文摘要

OBJECTIVES.: Several studies have demonstrated the benefits of tonotopic fitting on speech-in-noise perception in new cochlear implant (CI) users. However, these studies used different coding strategies (CIS, HDCIS, FSP, FS4, or mixed) with different frequency spectra. The aim of this study was to assess the impact of fine structure (FS) cues on speech perception in noise and in quiet and music perception, in new CI users with a tonotopic fitting map.
METHODS.: A prospective, randomized, double-blind, 2-period crossover study was performed from April 2023 to June 2024. All participants had tonotopic fitting determined by using flat-panel computed tomography (CT) and reconstruction software based on the Greenwood function. 24 new CI users aged 18 years or older with either bilateral severe-to-profound sensorineural hearing loss or complete hearing loss for less than 5 years were recruited at a single tertiary referral center. Participants were randomized to FS followed by conventional coding, or the reverse sequence, for 6-week periods with speech and music assessments after each period.
RESULTS.: FS cues significantly improved speech recognition in noise for high signal-to-noise ratio (SNR = +9 dB, Mean Effect (ME) = 6.0%, Standardized effect size (SES) = 0.62, p = 0.010). A period effect was observed beyond + 6 dB of SNR (SNR = +9 dB, p = 0.010; SNR = +6 dB, p = 0.018) without a statistically significant group effect. Speech recognition and tonal audiometry in quiet were not affected by FS coding. Melodic contour identification was enhanced by FS cues (ME = 7.2%, SES = 0.51, p = 0.030), and, in an exploratory subset analysis, low-frequency pitch discrimination was also improved (ME = 8.5 Hz, SES = 0.72, p = 0.04). Over 75% of the participants retained the tonotopic map with FS coding after study completion.
CONCLUSION.: FS coding strategies on tonotopic maps may improve speech recognition in noise at favorable SNRs, without compromising speech understanding in quiet. It may also enhance selected complex sound perception aspects, particularly pitch- and contour-related processing, while benefits for subjective music perception appear less consistent.