Monitoring senior wellness status using multimodal biosensors

Abstract

Because of the increased lifespan, there is an increasing demand on prevention of disease for senior wellness. For monitoring senior wellness status, biosensors such as Electroencephalography (EEG), Electrocardiography (ECG), blood pressure (BP), and respiration rate (RR) sensors and environmental sensors (temperature, humidity, motion, and light sensors) were used for data collection. Sensing data from bio- and environmental sensors are transferred to gateway in smart home and gateway send data into smart home server for the storage and analysis. Sensing data is analyzed using SVM for filtering, decision tree for health risk ratio, and EM algorithm for decision making. In this paper, we develop an EM-based inspection service middleware for monitoring elderly wellness status based on time and zone transition. This inspection service middleware for the prediction of abnormal health status has three steps as follows; monitoring, activity assessment, health risk assessment and decision-making. The activity assessment step used fuzzy logic, the health risk assessment step uses the decision-tree model for the classification of health status. Finally, EM-based decision-making step recommend exercise and housework for healthy status and rest or hospital checkup for tired/sick status. © 2016 IEEE.