A Novel Earphone Type Sensor for Measuring Mealtime: Consideration of the Method to Distinguish between Running and Meals.
A Novel Earphone Type Sensor for Measuring Mealtime: Consideration of the Method to Distinguish between Running and Meals.
复制标题
DOI：
10.3390/s17020252
复制
发表时间：
2017-01-27
期刊：
影响因子：
--
通讯作者：
Nishikawa A
中科院分区：
文献类型：
--
作者：
Taniguchi K;Chiaki H;Kurosawa M;Nishikawa A
在这项研究中，我们描述了一种技术，估计用餐时间使用耳机型可穿戴传感器。将由发光二极管和光电晶体管组成的小型光学传感器插入用户的耳孔中，并根据接收到的光量的时间变化来估计用户的进餐时间。这是通过向耳道内部发射光并接收从耳道反射回来的光来实现的。这项技术可以将“吃饭”与谈话、打喷嚏、走路、上下楼梯、操作电脑和使用智能手机区分开来。佩戴在用户头部上并测量食物摄入量的常规设备在跑步期间会振动，因为身体比步行期间受到更剧烈的震动;这可能导致这些设备将跑步误认为进食。为了解决这个问题，我们同时使用了两个传感器：一个在左耳，一个在右耳。这是基于我们的发现，即在跑步过程中，左右耳道的测量结果具有很强的相关性，但在进食过程中没有相关性。这允许基于相关系数来区分跑步和进食，这可以减少错误识别。此外，通过使用由半导体构成的光学传感器，可以制造小型且重量轻的装置。这种测量技术还可以测量与跑步相关的身体运动，从插入耳朵的光学传感器获得的数据可以用于支持有关饮食和运动的健康生活方式。
In this study, we describe a technique for estimating meal times using an earphone-type wearable sensor. A small optical sensor composed of a light-emitting diode and phototransistor is inserted into the ear hole of a user and estimates the meal times of the user from the time variations in the amount of light received. This is achieved by emitting light toward the inside of the ear canal and receiving light reflected back from the ear canal. This proposed technique allowed “meals” to be differentiated from having conversations, sneezing, walking, ascending and descending stairs, operating a computer, and using a smartphone. Conventional devices worn on the head of users and that measure food intake can vibrate during running as the body is jolted more violently than during walking; this can result in the misidentification of running as eating by these devices. To solve this problem, we used two of our sensors simultaneously: one in the left ear and one in the right ear. This was based on our finding that measurements from the left and right ear canals have a strong correlation during running but no correlation during eating. This allows running and eating to be distinguished based on correlation coefficients, which can reduce misidentification. Moreover, by using an optical sensor composed of a semiconductor, a small and lightweight device can be created. This measurement technique can also measure body motion associated with running, and the data obtained from the optical sensor inserted into the ear can be used to support a healthy lifestyle regarding both eating and exercise.