A Light-Weight Data Preprocessing and Integrative Scheduling Framework for Health Monitoring

The aging of population presents a severe challenge to the traditional "face-to-face" diagnosis mode. Household health monitoring systems play a more and more important role in identifying health conditions, diagnosing diseases, assisting daily lives, etc. However, most existing health monitoring systems are incomplete. They always leave out some function, e.g., data preprocessing and scheduling mechanism, which decrease the accuracy and efficiency of health services. In this paper, we propose a multi-function health monitoring system which integrates data acquisition, wireless data transmission, data storage, and data visualization. Moreover, we put forward a light-weight data preprocessing algorithm (LDPA) and an integrative scheduling mechanism (ISM) based on the amount of data and the semaphore. The LDPA can identify and eliminate abnormal data, while the ISM is able to significantly reduce the total time consuming. Experimental results based on 97 elderlies approved the high performance of LDPA in identifying abnormal breathing waves (the precision and the recall are 92.14% and 93.26% respectively) and the significant time reduction of ISM (saving 34.24% to 56.30% time overhead), compared with three other scheduling mechanisms.