新植入人工耳蜗使用者声感知精细结构声音编码策略的评估

文献信息

中文摘要

目的： 声音可分为两种时间信息形式：包络和精细结构。虽然人工耳蜗在安静环境下的言语识别中有效，但在噪声下聆听和音乐感知方面效果较差。本研究旨在比较精细结构处理策略（FS4）与既往编码策略（HDCIS）在新植入人工耳蜗使用者中的效果。
方法： 这项前瞻性、随机、交叉、双盲研究于2017年1月至2018年7月纳入19名参与者。每名参与者被随机分配接受先使用FS4编程的映射图、随后使用HDCIS编程的映射图，或相反顺序。在适应每种设置1个月后，进行安静环境下的纯音/言语测听、音乐感知测试（Gabrielsson量表、音程测试、旋律轮廓识别）以及生活质量（HISQUI）评估。
结果： 与HDCIS相比，FS4显著改善了安静环境下的言语识别（平均效应（ME）=8.4%，95%置信区间（CI95%）=3.5%至13.2%，p=0.004，标准化效应量（SES）=1.06）、Gabrielsson量表评分（ME=4.8，CI95%=0.4至9.1，p=0.047，SES=0.67）和音程感知（ME=18 Hz，CI95%=2.7至32.4 Hz，p=0.024，SES=0.78）。旋律轮廓识别或HISQUI未发现显著差异。在安静环境下的言语识别（p=0.002）、Gabrielsson量表（p=0.005）、音程测试（p=0.032）和HISQUI（p=0.039）中观察到时期效应。80%的参与者在研究期结束后保留了FS4设置。
结论： 在新植入人工耳蜗使用者中，与HDCIS相比，FS4与音乐感知和安静环境下言语识别的改善相关，提示精细结构处理策略可能具有优势。西V伦理委员会编号：2016-A01952-49。2016年12月28日。

英文摘要

PURPOSE: Sound can be divided into 2 forms of temporal information: envelope and fine structure. While cochlear implants are effective in speech recognition in quiet, they are less effective for listening in noise and for music perception. This study aimed to compare a fine structure processing strategy (FS4) with a previous coding strategy (HDCIS) in newly implanted CI users.
METHODS: This prospective, randomized, cross-over, double-blind study included 19 participants between January 2017 and July 2018. Each participant was randomized to receive a map programmed with FS4 followed by a map programmed with HDCIS or vice versa. Tonal/vocal audiometry in quiet, music perception tests (Gabrielsson scale, interval test, melodic contour identification), and quality of life (HISQUI) assessments were conducted after 1 month of adaptation to each setting.
RESULTS: FS4 significantly improved speech recognition in quiet (mean effect (ME) = 8.4%, 95% confidence interval (CI95%) = 3.5% to 13.2%, p = 0.004, standardized effect size (SES) = 1.06), Gabrielsson scale scores (ME = 4.8, CI95% = 0.4 to 9.1, p = 0.047, SES = 0.67), and interval perception (ME = 18 Hz, CI95% = 2.7 to 32.4 Hz, p = 0.024, SES=0.78). No significant difference was found for melodic contour identification or HISQUI. A period effect was observed for speech recognition in quiet (p = 0.002), Gabrielsson scale (p = 0.005), interval test (p = 0.032), and HISQUI (p = 0.039). Eighty percent of the participants kept the FS4 setting after the study period.
CONCLUSION: FS4 was associated with improved music perception and speech recognition in quiet compared to HDCIS in newly implanted cochlear implant users, suggesting potential advantages of fine structure processing strategies. N° West V Ethics Committee: 2016-A01952-49. December 28, 2016.