ACTIVE AND NONLINEAR MODELS FOR COCHLEAR MECHANICS

为了更好地理解听力过程，人们在耳蜗建模和生理测量方面做了大量的工作。了解正常听觉功能的过程对于帮助确定听力损失和耳鸣的原因具有重要意义。了解每个组件的形态和功能可以为开发更好的人工耳蜗和非侵入性确定疾病的诊断过程奠定基础。非听觉系统的语音识别和信号处理的生物灵感设计也是可能的，并可能对听力以外的应用产生重大影响。目前对耳蜗力学的研究主要集中在确定体内测量中所见的增强滤波和非线性压缩的来源。自1948年首次提出有源放大过程是增强滤波的来源以来，这一假说得到了广泛的研究。在这项工作中，提出了一种全面、有效的非线性和主动宏观耳蜗力学的数值策略。通过这一能力，将开发一个用于模型测试的虚拟实验室，能够纳入关于活动和几何不一致性的最一般的非线性模型。采用解析和数值混合的方法，将Corti器官的微观力学，特别是外毛细胞及其连接结构，包括在全局响应模型中。这些预测将与从生理实验中获得的体内数据进行比较。实验验证是这个建模工作的中心焦点。与生理测量的密切联系对于验证建模参数很重要，最重要的是验证假设的换能器模型与耳蜗内电位(或电流)的关系。受控实验将用于识别换能器模型参数(例如，增益)和验证/无效假设。

In order to better understand the hearing process, numerous efforts in cochlear modeling and physiological measurement have been undertaken. Understanding the process of normal auditory function holds the significant promise of assisting in the determination of the causes of hearing loss and tinnitus. Understanding the morphology and function of each component can lead the way to the development of better cochlear prostheses and diagnostic processes for noninvasive determination of disease. Biologically inspired designs for speech recognition and signal processing of non--auditory systems are also possible and could have a significant impact for applications other than hearing. Current research in cochlear mechanics is focussed on determining the source of the enhanced filtering and nonlinear compression seen in in vivo measurements. The hypothesis that an active amplification process is the source of the enhanced filtering has been studied widely since first proposed in 1948. In this grant a comprehensive, efficient numerical strategy for nonlinear and active macroscopic cochlear mechanics is proposed. Through this capability, a virtual laboratory for model testing will be developed capable of incorporating the most general nonlinear models for activity and geometric nonuniformity. Using a hybrid analytic and numeric approach the micromechanics of the Organ of Corti, especially the outer hair cells and their connecting structures, will be included in the global response modeling. These predictions will be compared to in vivo data obtained from physiological experiments. Experimental validation is a central focus of this modeling effort. Close ties to the physiological measurements is important to validate modeling parameters, most importantly the relation of the hypothesized transductions models to the endocochlear potential (or current). Controlled experiments will be used to both identify transducer model parameters (e.g., gains) and to validate/invalidate the hypothesis.