Modeling and analyzing the acoustic signatures of human swallowing mechanism

吞咽是人体最复杂的行为之一。即使吞咽事件的顺序稍有延迟或不匹配，也可能导致食物吸入肺部，并导致明显的发病率和死亡率。尤其是在老年人和残疾、中风、头颈或脊髓损伤的患者中，吞咽障碍相当常见。为了评估吞咽障碍，患者通常在进食过程中进行特殊的X光(视频荧光镜)观察。由于我们最近的研究表明吞咽声分析具有很好的诊断能力，我们建议在这里研究其他使用声学分析进行吞咽评估的方法。利用现有的食道压力测压数据与视频透视图像和吞咽声音相对应，我们将为吞咽机制开发一个全面的结构模型，并模拟特定吞咽事件的声学特征。我们将使用在健康受试者和患者中获得的数据来验证和测试我们的模型。我们假设，模型的一些参数，即那些代表肌肉力量或结构性运动之间的时间安排的参数，在两组之间将显著不同，从而提供了一种客观探索吞咽障碍原因的方法。我们还建议开发一个集成和模块化系统的原型，以采集、去噪、分析和筛选吞咽和呼吸声音信号，并使用无线技术将数据传输到远程临床诊断中心。

Swallowing is among the most complicated acts of the human body. Even a slight delay or mismatch in the sequence of swallowing events may allow food aspiration into the lung and causes appreciable morbidity and mortality. Especially in the elderly and in patients with disabilities, stroke, head-neck or spinal cord injuries, swallowing disorders are quite common. To assess swallowing disorders, patients are generally observed with a special X-ray (videofluroscopy) during feeding. Since our recent studies have shown promising diagnostic capabilities of swallowing sound analysis, we propose here to investigate other methods for swallowing assessment using acoustical analysis. Using available manometry data of the esophageal pressure in correspondence to videoflouroscopic images and swallowing sounds, we will develop a comprehensive structural model for the swallowing mechanism and model the acoustic signatures of specific swallowing events. We will validate and test our model with data obtained in both healthy subjects and patients. We hypothesize that some of the model's parameters, i.e. those that represent the muscle force or the timing between structural movements, will be significantly different between the two groups, and hence provide a means to explore objectively the cause of swallowing disorders. We also propose to develop a prototype of an integrated and modular system to acquire, denoise, analyze, and screen swallowing and breath sound signals, and to use wireless technology to transfer data to a remote clinical diagnostic center.