Detection of chewing and swallowing to estimate eating patterns and energy intake

描述(由申请人提供)： 发达国家的肥胖率正在以惊人的速度增长。肥胖会增加患心脏病、高血压、糖尿病和某些癌症的风险，现在被认为是心血管疾病的风险因素。这项研究的目的是探索新型非侵入性设备和模式识别方法的应用，以进行人类食物摄取行为的研究，并产生对容量和热量食物摄入量的客观估计，这将有助于确定有效的措施来治疗或预防肥胖等疾病。这样的设备和方法可以扩大我们对肥胖原因的理解，这项研究中创建的监测设备可以用于监测肥胖患者，并作为治疗的一部分，潜在地改善生活质量，降低与肥胖相关的发病率和死亡率。这项研究的目的是设计和执行一项初步调查，提供初步数据，即通过可穿戴的非侵入性监测设备对咀嚼(咀嚼)和吞咽(吞咽)进行客观观察，通过提供具有足够灵敏度和特异度的妄想频率、咀嚼持续时间和确定食物摄入时间的客观测量，提供统计上可靠的饮食习惯估计。 拟议研究的目标包括设计可穿戴传感器和相关的信号处理方法；开发模式识别方法，从传感器记录中自动检测错觉和咀嚼的情况；开发模式识别方法，根据检测到的咀嚼和吞咽来自动识别食物摄入的时间段；并在一组人类受试者身上验证这些设备和方法。现代计算智能方法，如人工神经网络和模糊逻辑，将与统计方法一起使用，以实现最高的模式识别精度。

DESCRIPTION (provided by applicant): The prevalence of obesity in developed countries is increasing at an alarming rate. Obesity contributes to an increased risk of heart disease, hypertension, diabetes, and some cancers and is now considered a risk factor for cardiovascular disease. The objective of this research is to investigate application of novel noninvasive devices and pattern recognition methods to perform studies of human food intake behavior and produce objective estimates of volumetric and caloric food intake that will be relevant for identifying effective measures to treat or prevent diseases like obesity. Such devices and methods could extend our understanding of causes of obesity, and the monitoring devices created in this study could be used for monitoring of obese patients and as a part of a therapy potentially improve quality of life and decrease the morbidity and mortality associated with obesity. The goal of this study is to design and perform a pilot investigation which will provide preliminary data that objective observations of mastication (chewing) and deglutition (swallowing) by a wearable, non-intrusive monitoring device can provide statistically reliable estimates of eating habits by providing objective measures of delution frequency, duration of mastication and identifying periods of food intake with sufficient sensitivity and specificity. The aims of the proposed research include design of the wearable sensors and associated signal processing methods; development of pattern recognition methodologies that will automatically detect instances of delution and mastication from sensor recordings; development of pattern recognition methods to automatically identify periods of food intake based on detected chewing and swallowing; and to validate these device and methodologies on a group of human subjects. Modern methods of computational intelligence such as artificial neural networks and fuzzy logic will be used along with statistical methods to achieve the highest 'accuracy of pattern recognition.