PDhms: Pulse Diagnosis via Wearable Healthcare Sensor Network

Pulse Diagnosis Theory (PDT) has the advantages of non-invasive treatment and disease prevention. Combining these merits with Wireless Sensor Network (WSN), we propose a novel networked low-cost and wearable healthcare monitoring system, namely PDhms, for pulse data collection, pulse analysis and pulse diagnosis. Some practical challenges still exist in PDhms such as exerting appropriate pressure on a human radial artery, overcoming seriously limited resources, improving low signal to noise ratio (SNR) and conducting resistance of interference. To address these challenges, we present a robust external pressure control algorithm for pulse data collection, and propose FEA, a novel light-weight and adaptive feature extraction algorithm for sensed pulse data. We conduct the large-scale pulse data collection experiments of 1356 pulse samples, the comparison experiment between the FEA and the typical derivative-based algorithm, as well as pulse diagnosis experiments based on support vector machine (SVM). Experimental results show that PDhms is a valuable solution for low-cost wearable healthcare monitoring system.