阈上听觉功能的特征描述：22项直接与派生听觉测量指标的比较

文献信息

中文摘要

背景： 相当一部分人群的临床听阈正常，却在背景噪声中理解言语存在困难。现有听力模型使用有限数量的生理和信号处理机制来解释这些困难，这意味着心理声学测量指标之间的关系可能反映共同的基础过程。然而，这些关系很少在同一批受试者中被全面检验。本研究使用线性回归和主成分分析量化心理声学测量指标之间的关联。方法：在52名年龄在20-80岁、具有年龄典型或轻度听力损失（PTA4 < 32.5 dB HL）的志愿者中，获取并分析单耳和双耳聆听条件下的噪声中言语、不同噪声类型中的噪声中纯音以及响度增长函数。结果：线性回归分析确定直接心理声学测量指标——尤其是耳间反相噪声中纯音（N0Sπ）阈值——是阈上听觉功能的稳健指标。尽管N0Sπ与相应的同相双耳阈值（N0S0）仅呈中等关联，但其与较差耳噪声中言语阈值的关联（r = 0.60）强于相应的500 Hz听阈（r = 0.35），提示该关系并非仅由可听度驱动。探索性主成分分析进一步显示，N0Sπ阈值对与噪声中言语变异相关的多个听觉维度均有贡献。相比之下，双耳掩蔽级差或听觉滤波器带宽等派生指标显示出较弱且较不一致的关联。结论：总体而言，结果表明选定的直接心理声学测量指标可能比派生指标更稳健地表征阈上听觉功能。

英文摘要

A substantial proportion of the population has clinically normal hearing thresholds yet experiences difficulty understanding speech in background noise. Existing hearing models explain these difficulties using a limited number of physiological and signal-processing mechanisms, implying that relationships among psychoacoustic measures may reflect shared underlying processes. However, these relationships have rarely been examined comprehensively within the same participants. Here we quantified associations among psychoacoustic measures using linear regression and principal component analysis. Speech-in-noise, tone-in-noise in different noise types, and loudness growth functions were obtained and analyzed in monaural and binaural listening conditions in 52 volunteers (aged 20-80 years) with age-typical or mild hearing loss (PTA4 < 32.5 dB HL). Linear regression analyses identified direct psychoacoustic measures - particularly interaurally antiphasic tone-in-noise (N0Sπ) thresholds - as robust indicators of suprathreshold hearing function. Although N0Sπ showed only a moderate association with the corresponding diotic thresholds (N0S0), it was more strongly associated with poorer-ear speech-in-noise thresholds (r = 0.60) than the corresponding 500-Hz audiometric threshold (r = 0.35), suggesting that this relationship was not driven by audibility alone. Exploratory principal component analysis further showed that N0Sπ thresholds contributed to multiple auditory dimensions associated with speech-in-noise variability. In contrast, derived measures such as binaural masking level difference or auditory filter bandwidth showed weaker and less consistent associations. Overall, the results indicate that selected direct psychoacoustic measures may provide a more robust characterization of suprathreshold auditory function than derived measures.