Development of a triage engine enabling behavior recognition and lethal arrhythmia detection for remote health care system

• Published in: 2011 Annual International Conference of the IEEE Engineering in Medicine and Biology Society Vol. 2011; pp. 2160 - 2163
• Main Authors: Sugano, H., Hara, S., Tsujioka, T., Inoue, T., Nakajima, S., Kozaki, T., Namkamura, H., Takeuchi, K.
• Format: Conference Proceeding Journal Article
• Language: English
• Published: United States IEEE 01.01.2011
• Subjects: Actigraphy - instrumentation; Arrhythmias, Cardiac - diagnosis; Arrhythmias, Cardiac - mortality; Death, Sudden, Cardiac - prevention & control; Diagnosis, Computer-Assisted - instrumentation; Diagnosis, Computer-Assisted - methods; Electrocardiography; Engines; Equipment Design; Equipment Failure Analysis; Humans; Medical services; Receivers; Reproducibility of Results; Sensitivity and Specificity; Telemedicine - instrumentation; Telemedicine - methods; Telemetry - instrumentation; Telemetry - methods; Three dimensional displays; Triage - methods; User-Computer Interface; Vital Signs; Wireless communication; Wireless sensor networks
• ISBN: 9781424441211; 1424441218
• ISSN: 1094-687X; 1557-170X
• DOI: 10.1109/IEMBS.2011.6090405

For ubiquitous health care systems which continuously monitor a person's vital signs such as electrocardiogram (ECG), body surface temperature and three-dimensional (3D) acceleration by wireless, it is important to accurately detect the occurrence of an abnormal event in the data and immediately inform a medical doctor of its detail. In this paper, we introduce a remote health care system, which is composed of a wireless vital sensor, multiple receivers and a triage engine installed in a desktop personal computer (PC). The middleware installed in the receiver, which was developed in C++, supports reliable data handling of vital data to the ethernet port. On the other hand, the human interface of the triage engine, which was developed in JAVA, shows graphics on his/her ECG data, 3D acceleration data, body surface temperature data and behavior status in the display of the desktop PC and sends an urgent e-mail containing the display data to a pre-registered medical doctor when it detects the occurrence of an abnormal event. In the triage engine, the lethal arrhythmia detection algorithm based on short time Fourier transform (STFT) analysis can achieve 100 % sensitivity and 99.99 % specificity, and the behavior recognition algorithm based on the combination of the nearest neighbor method and the Naive Bayes method can achieve more than 71 % classification accuracy.