Noise, stress, and the perception of loudness: exploring the mechanistic bases of tinnitus and hyperacusis with the use of technological advances

耳鸣在大多数情况下与一生中听力功能的自然下降有关。最近的数据显示，它在年轻群体中的流行率呈上升趋势。听觉亢进，对外界声音的敏感性增加，可能与耳鸣有共同的机制，包括中央增益功能上调。值得注意的是，这两种情况似乎都因个体的压力状态而加剧。然而，在正常听力的成年人中，噪音和压力的相互作用知之甚少。因此，我们建议研究对噪音和压力的动态反应，表征i)与噪音相关的自主心脏反应，身体和心理压力（第1部分），以及ii)解决与不同面部表情模式相关的响度的感觉和情感维度（第2部分）。在第一部分中，我们建议将听觉研究平台（ARP）从一个听力保护装置转变为一个强大的研究仪器，利用位于闭塞的听道内的内部麦克风来提取有节奏的心脏信号，以评估对噪音和压力的自主反应，并检验关于男性和女性以及年轻人和老年人差异的假设。来自ARP声学记录的心跳检测将用于从暴露于几个实验条件（~5分钟持续时间）的正常听力成人中获得正常心跳间隔（NN）的时间序列。然后对数据进行心率变异性（HRV）分析，以确定高（高频，0.15-0.4 Hz）和低频（低频，0.04-0.15 Hz）频谱成分的功率，作为副交感神经反应和交感/副交感神经平衡变化的测量，运行ANCOVAs（性别和年龄作为受试者之间和条件变量，以及问卷得分作为协变量）对HR和HRV指数进行分析。在第二部分中，我们将研究正常听力成年人在听到不同感官（强度）和情感维度（厌恶与中性或愉快）的声音时，会表现出特定的面部表情模式的假设。因此，我们将记录参与者的动态面部表情。与其使用经典的手动且耗时的面部动作编码系统（FACS），我们建议使用一种机器学习方法，该方法可以快速处理具有异构分布的大型复杂数据，同时-与经典FACS相反-允许识别新模式来预测特定分类器，这里的感官和情感响度评级，作为每个维度的最相关指标。这些研究解决了对噪音和压力的动态生理反应，这两者在我们的当代环境中都在上升，使用新颖的方法，有望在现实生活的动态环境中提供对听力障碍的机制基础的见解。

Tinnitus is associated in most cases with a natural decline in hearing function over a lifetime. Recent data show that its prevalence is on the rise among younger age groups. Hyperacusis, an increased sensitivity to external sounds, possibly shares with tinnitus common mechanisms including upregulated central gain function. Remarkably, both appear to be exacerbated by the individual's stress status. However, little is known regarding noise and stress interactions in normal hearing adults. We therefore propose to investigate dynamic responses to noise and stress, characterizing i) autonomic cardiac responses to noise-related, physical and psychological stress (Part 1), and ii) addressing the sensory and affective dimensions of loudness in relation to differential facial expression patterns with which they are putatively associated (Part 2). In Part 1, we propose to transform the Auditory Research Platform (ARP)'s use from a hearing protection device into a powerful research instrument taking advantage of the internal microphone located in the occluded auditory canal to extract rhythmic cardiac signals for assessment of autonomic responses to noise and stress and test hypotheses concerning differences between male and female as well as young and aging adults. Heart beat detection from ARP acoustic recordings will be used to derive times series of intervals between normal beats (NN) from normal-hearing adults exposed to several experimental conditions (~5 min duration). The data will then be subjected to heart rate variability (HRV) analysis to determine power for the high (HF, 0.15-0.4 Hz) and low frequency (LF, 0.04-0.15 Hz) spectral components as measures of parasympathetic responses and changes in sympathetic/ parasympathetic balance, running ANCOVAs (Sex and Age as between- and Conditions as within- subjects variables, and questionnaire scores as co-variates) on HR and HRV indices. In Part 2, we will investigate the hypothesis that normal hearing adults, when exposed to sounds varying in their sensory (intensity) and affective dimensions (aversive versus neutral or pleasant), display specific patterns of facial expressions. We will therefore record dynamic facial expressions of participants when. Rather than using the classical manual and time-consuming Facial Action Coding System (FACS), we propose to make use of a machine-learning approach which can rapidly handle large bodies of complex data with heterogeneous distributions while also - in contrast to classical FACS - allowing the identification of novel patterns to predict specific classifiers, here sensory and affective loudness ratings, as the most relevant indices for each of these dimensions. These studies address dynamic physiological responses to noise and stress, both on the rise in our contemporary environment, using novel approaches that hold the promise of providing - in a real-life ambulatory context - insights into the mechanistic bases of hearing disturbances.