A feasibility study of health monitoring system for elederly people living alone

点击翻译按钮获取中文摘要

Purpose : To develop a remote health monitoring system for eldely people living alone which can decrease anxiety not only for the elderly but also his/her family.Attention point : Long-term monitoring is needed to estimate the health condition by measurement results. Then monitoring should be wholly unobtrusive.Results : In order to achieve the purpose mentioned above, we have gotten the following results.(1) Human motion detecting sensor : We decrease the volume of a sensor 1/3, and the weight of it 1/2 by using a new type of pyro-electric infra-red sensing device. We realized simple wiring and cost cutting.(2) In-house unit : We decrease the volume and the weight of the unit 1/10 by unifying the information collection and transfer function. We realized extremly large cost cutting.(3) Database server : We have developed a database system which enables us to whole measurement results seamlessfully and multi-dimensionaly. Adding to that, we achieved a function to send information of the elderly to his/her family via Internet using effective displaying methods.(4) Statistical analysing methods : We increased the practicability and accuracy of our statistical analysing methods of the monitored data using transition matrices or diagrams. Each matrix or graph usually has many similarities but it is difficult to prove that each daily pattern is also similar because the beginning time or continuing period are not the same. We solved these problems using the Dynamic Programming matching method. Adding to that, we are testing the practicability of an automatic emergency call function by checking staying period in a specified room or total length of a no response period.(5) Accelerometer : We also developed some portable movement sensors using 3-dimensional accelerometers to measure the quantity of daily exercise of the elderly. We expect that this sensor will supplement the infra-red sensor because its intimacy to the human body can be useful for detecting falls.