Developing a clinical diagnostic tool for age-related cochlear synaptopathy.

年龄相关性听力损失（ARHL）是主要的老年神经退行性疾病。最近的动物研究表明，最早由噪音和衰老引起的耳蜗病理包括内毛细胞带状突触的丧失。这种突触丧失的早期后果被认为不会影响听力灵敏度，而是会损害背景噪音下的言语理解。虽然这种被称为耳蜗突触病或隐藏性听力损失的情况可以通过听觉脑干反应（ABR）的阈上声诱发波-1振幅的减少在动物模型中检测到，并通过死后定量组织学证实，但其在人类中的诊断仍然具有挑战性。几种非侵入性电生理方法如ABR和耳蜗电图（ECochG）已被用于检测人类耳蜗突触病。他们提出的应用的两个主要障碍包括：(1)ABR/ECochG波振幅的高度可变性；(2)缺乏对活人的直接组织学验证。我的策略是开发一种新的电气设备，能够产生校准脉冲，以减少ABR/ECochG波指标的高可变性。验证问题将通过应用机器学习首先在动物模型中识别与耳蜗突触病变相关的多个ABR/ECochG标记来解决，其中耳蜗突触病变可以通过组织学突触计数直接验证，然后将这些识别的特征应用于人类诊断。我的假设是，通过改进与ABR/ECochG数据收集和分析相关的硬件和软件，可以实现人类耳蜗突触病的临床诊断。基于我们的初步研究，我们将继续改进我们的电路，以测试我们的校准脉冲装置是否可以解决变异性问题，同时，我们将在小鼠中识别与年龄相关的耳蜗突触病相关的ABR/ECochG特征，然后在沙鼠中验证这些识别的特征，因为沙鼠的听力范围与人类相似。最后，我将进行一项纵向研究，以测试人类在衰老过程中确定的特征。早期发现耳蜗突触病变可能会导致有效的预防策略，从而延迟或潜在地阻止ARHL的进一步发展。因此，该项目的完成预计将导致当前ARHL临床诊断的重大转变，同时为首席研究员提供一个独特的培训机会，以发展开发商业医疗设备的技能。

Age-related hearing loss (ARHL) is the predominant neurodegenerative disease of aging. Recent animal studies have demonstrated that the earliest cochlear pathology due to noise and aging involves a loss of the inner hair cell ribbon synapse. The early consequences of this synaptic loss are not believed to affect hearing sensitivity, but rather, impair speech understanding in background noise. While this condition, called cochlear synaptopathy or hidden hearing loss, can be detected in animal models by a reduction of suprathreshold sound evoked wave-1 amplitude of auditory brainstem response (ABR) and confirmed by post-mortem quantitative histology, its diagnosis in humans remains challenging. Several non-invasive electrophysiological methods such as ABR and electrocochleography (ECochG) have been tested to detect human cochlear synaptopathy. Two major obstacles to their proposed application include: (1) a high variability of ABR/ECochG wave amplitudes, and (2) a lack of direct histological validation for living humans. My strategy here is to develop a new electrical device able to generate a calibration pulse to reduce high variability in ABR/ECochG wave metrics. The validation issue will be addressed by applying machine learning to identify multiple ABR/ECochG markers associated with cochlear synaptopathy first in animal models, in which cochlear synaptopathy can be directly validated by histologic synaptic counting, and then apply these identified features for human diagnosis. My hypothesis is that clinical diagnosis of human cochlear synaptopathy can be achieved by improving both hardware and software related to ABR/ECochG data collection and analysis. Based on our preliminary studies, we will continue to improve our electrical circuit to test if our calibration pulse device can address the variability issue, and at the same time, we will identify ABR/ECochG features associated with age-related cochlear synaptopathy in mice and then validate these identified features in gerbils since its hearing range is similar to humans. Finally, I will perform a longitudinal study to test identified features in humans during aging. Early detection of cochlear synaptopathy may lead to effective preventive strategies that would delay or potentially prevent further development of ARHL. Completion of this project is therefore expected to lead to a major shift in current clinical diagnosis of ARHL, while providing a unique training opportunity for the principal investigator to develop skills in developing commercial medical devices.