AUDITORY PROCESSING OF COMPLEX SOUNDS

描述(申请人提供)：如果没有对声音的生理反应的详细知识，就不能理解听觉的神经机制，其中心理物理反应被很好地描述。这项建议提出了一个全面的方法来解决这个重要的问题，并集中在复杂声音的幅度调制所携带的信息如何被大脑编码和处理的问题上。尽管数字助听器技术最近取得了进展，但我们对处理复杂声音所涉及的神经机制的有限理解仍然严重限制了我们帮助听力损失的听者的能力。在我们对健康的听觉系统如何应对这一挑战的调查中，将出现帮助听众在复杂的声学环境中处理声音的新策略。这项建议的三个目标是行为、生理和计算建模方法的新组合，以解决幅度调制(AM)声音的编码和处理问题。第一个目标将测试关于行为和心理物理阈值的三个假设。第一个假设侧重于定义兔和人的AM检测和辨别阈值，并使用与生理方法兼容的严格匹配的测试程序(目标2)。第二个假设探索了一个长期存在的谜题：行为AM检测阈值随着声级的增加而提高，而单个单位的生理编码(基于当前理论)随着声级的增加而下降。第三个假设涉及到对掩蔽AM检测的详细线索的识别。这些提示将通过调制域中可再现掩码的新应用来识别。第二个目的是在清醒的兔下丘(IC)水平上检验生理性AM编码的假说。这些研究将包括根据目标1的行为结果选择刺激。我们开发了新的生理学方法，使用四极管从IC神经元群体中进行时间上的精确记录。这些记录能够严格测试神经编码对平均放电率和时间反应模式的相对依赖，包括对神经元集合的尖峰率和模式的统计分析。第三个目标使用计算技术来测试IC中AM-Rate调谐的竞争理论。最近的模型提出了不同的神经机制来解释中脑的AM反应。我们将严格测试这些模型，并将包括使用模型设计所针对的刺激以外的测试。结果将明确决定哪些模型与我们的生理数据最一致。这些研究将促进我们对听觉系统中AM编码和处理的潜在机制的理解。这些信息将指导通过改进助听器信号处理算法来增强和恢复听力损失听众复杂声音的关键方面的努力。这个项目的公共卫生相关性是确定健康的听觉系统如何编码复杂的声音。我们将使用行为、生理和计算机建模方法的新组合来识别大脑如何编码和提取复杂声音中的幅度波动。由于人类的听力损失通常涉及理解复杂声音的困难，了解大脑如何编码这些无处不在的声音将为帮助听力损失的听者提供新的重要见解。

DESCRIPTION (provided by applicant): The neural mechanisms of auditory perception cannot be understood without detailed knowledge of physiological responses to sounds for which psychophysical responses are well described. This proposal presents a comprehensive approach to this important problem and focuses on the question of how information carried by amplitude modulations of complex sounds is encoded and processed by the brain. Despite recent advances in digital hearing-aid technology, our limited understanding of the neural mechanisms involved in processing complex sounds remains a significant limitation in our ability to aid listeners with hearing loss. New strategies to assist listeners with processing sounds in complex acoustic environments will emerge from our investigations of how the healthy auditory system handles this challenge. The three Aims of this proposal feature a novel combination of behavioral, physiological, and computational modeling approaches to address the problem of encoding and processing amplitude-modulated (AM) sounds. The 1st Aim will test three hypotheses concerning behavioral and psychophysical thresholds. The first hypothesis focuses on defining AM detection and discrimination thresholds for rabbits and humans and uses rigorously matched test procedures that are compatible with physiological approaches (Aim 2). The second hypothesis probes a long-standing puzzle: behavioral AM-detection thresholds improve as sound level increases, whereas single-unit physiological coding (based on current theories) degrades as level increases. The third hypothesis concerns the identification of detailed cues for masked AM detection. These cues will be identified with a novel application of reproducible maskers in the modulation domain. The 2nd Aim will test hypotheses of physiological AM coding at the level of the inferior colliculus (IC) in awake rabbit. These studies will include stimuli selected on the basis of behavioral results from Aim 1. We have developed new physiological methods for temporally precise recordings from populations of IC neurons using tetrodes. These recordings enable rigorous tests of the relative reliance of neural encoding on average discharge rates and temporal response patterns, including statistical analyses of spike rates and patterns across ensembles of neurons. The 3rd Aim uses computational techniques to test competing theories for AM-rate tuning in the IC. Recent models have proposed various neural mechanisms to explain AM responses in the midbrain. We will rigorously test these models and will include tests with stimuli other than those for which the models were designed. The results will explicitly determine which models are most consistent with our physiological data. These studies will advance our understanding of the mechanisms underlying AM coding and processing in the auditory system. This information will instruct efforts to enhance and restore critical aspects of complex sounds for listeners with hearing loss by improving hearing-aid signal-processing algorithms. The Public Health Relevance of this project is to determine how the healthy auditory system encodes complex sounds. We will use a novel combination of behavioral, physiological, and computer modeling approaches to identify how the brain encodes and extracts amplitude fluctuations in complex sounds. Because hearing loss in humans typically involves difficulty understanding complex sounds, knowledge of how the brain codes these ubiquitous sounds will provide new and important insights for aiding listeners with hearing loss.