早期频谱分辨率可预测成人人工耳蜗植入者后期言语识别
这项前瞻性纵向研究纳入531例语后聋成人人工耳蜗植入者,从开机到术后1年评估频谱分辨率(EasyQSMD)和言语识别(CNC、AzBio)。模型推导的早期EasyQSMD阈值可显著预测后期言语识别,解释3%-8%的方差;频谱分辨率在开机后1周即达平台期。
要点
- 频谱分辨率在开机后1周内快速稳定并维持至1年
- 早期EasyQSMD阈值可显著预测6个月和1年言语识别
- EasyQSMD简短且不依赖语言内容,适合多语言人群早期评估
核心结论
- 成人人工耳蜗植入者频谱分辨率在开机后1周达平台期,早期EasyQSMD阈值对后期言语识别有显著预测价值,可解释3%-8%的方差。
研究设计与样本
- 前瞻性纵向研究,纳入531例语后聋成人人工耳蜗植入者,从开机随访至术后1年。
- 采用生长曲线模型拟合各时间点EasyQSMD阈值,生成开机后1个月的早期频谱分辨率模型估计值。
- 通过线性回归分析,用模型推导的早期阈值预测6个月和1年(取最佳值)的言语识别结果。
主要发现
- EasyQSMD阈值从开机到开机后1周改善,随后至第一年保持稳定。
- 模型推导的早期EasyQSMD阈值显著预测后期言语识别表现,解释3%-8%的方差。
关键数值与推荐
- 言语识别评估包括CNC单词和AzBio句子(安静与噪声环境)。
- 频谱分辨率评估工具为EasyQSMD。
- 预测因素为模型推导的开机后1个月EasyQSMD阈值。
临床怎么用
- 可将频谱分辨率评估纳入开机后早期临床方案,用于个体化咨询和设定现实预期。
- 早期识别表现不佳者,可在言语识别技能尚在发展、神经可塑性高峰的关键窗口进行靶向干预。
- EasyQSMD简短且不依赖语言内容,适用于多语言人群及耳蜗植入后早期听觉康复阶段。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
OBJECTIVES: To investigate the trajectory of spectral resolution in adult cochlear implant (CI) recipients and determine whether early spectral resolution measures can predict later speech recognition outcomes. METHODS: Prospective, longitudinal study of 531 postlingually deafened adult CI recipients. Spectral resolution (EasyQSMD) and speech recognition (CNC words, AzBio sentences in quiet and noise) were evaluated from activation through 1-year post-activation. Growth curve models were fit to EasyQSMD thresholds across all timepoints to generate mode-based estimates of early spectral resolution at 1-month post-activation. These model-derived early thresholds were then used to predict later speech recognition outcomes (best of 6-months and 1-year) through linear regression analyses. RESULTS: EasyQSMD thresholds improved from activation to 1-week post-activation, then stabilized through the first year. Model-derived early EasyQSMD thresholds significantly predicted later speech recognition outcomes, explaining 3%-8% of the variance in performance. CONCLUSION: Spectral resolution rapidly stabilizes, reaching asymptotic performance by 1-week post-activation. The EasyQSMD provides a non-linguistic assessment tool that demonstrates significant predictive value for subsequent speech recognition outcomes. This early predictive capability enables the identification of poor performers during a critical window when speech recognition skills are still developing and the brain's neuroplasticity is at peak potential. Integrating spectral resolution assessment into early post-activation clinical protocols could improve individualized realistic expectations counseling and guide targeted interventions-potentially leading to improved overall CI adoption rates. The brief administration time and independence from linguistic content make the EasyQSMD particularly valuable for multilingual populations and during the earliest stages of auditory rehabilitation post-CI.
引用
- Berg KA, Roberts JB, Burchesky MZ, et al. Early Spectral Resolution Predicts Later Speech Recognition in Adult Cochlear Implant Recipients. Laryngoscope. 2026. PMID: 41730806. doi:10.1002/lary.70383
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。