무선 휴대용 심전도 모니터링 시스템 구현에 관한 연구

Abstract

The electrocardiogram (ECG) is a representative signal containing information about the condition of the heart. However, the heart diseases may not show up all the time, but would manifest at certain irregular intervals during the day. Therefore human observer wants to get the ECG data by ambulatory ECG monitoring system. The existing ECG monitoring system has a lot of constraints for mobile environments, especially on low power consumption and irregular outdoor environment, etc. In this study, we implemented the 3 channel wireless ECG monitoring system. This system consists of 3 wireless sensor nodes with ECG sensor board(including 3 electroid per each sensor node), 1 base node for receiving the ECG data transmission and 1 ECG monitoring software for display on PC. We used PowerTOSSIM to estimate power consumption of wireless ECG monitoring system. From the experiments, we found that radio spends most power since radio works continuously for data transmission and we can't change it into sleeping mode. We adopted the recent study of "Intelligent sensor node based a low power ECG monitoring system" by Minzeng et al to reduce radio power consumption. The node equipped with intelligence can decide whether the captured signal is normal or abnormal by following the procedure: A cardiac cycle is captured with 300 hundred samples. After that, euclidean distance is calculated according to AR modeling and classify abnormal ECG cycles. If ECG sensor node determines that the signal is abnormal, the transmission will be launched. Sensor node transmits only abnormal ECG cycles to the base node and will not transmit any data with normal classification. From the experiments, we show that we can save maximum 57% power consumption by adopting the intelligent sensor nodes.