Publications

Abstract : Remote health monitoring and intervention systems including wearable sensors, smartphones and advanced communication technologies are slated to be a game changer in the delivery of quality healthcare services, especially in developing parts of the world. However, we are yet to see large scale adoption of remote health monitoring systems due to many factors such as: lack of reliable data network coverage, high power requirements for smartphone analytics, and unreliability in the timely delivery of critical data to remote doctors. In addition to these, the huge volume of sensor data and alerts from multiple remote patients are unmanageable for already overloaded doctors. In this paper, we attempt to address each of these issues. First, we propose a novel healthcare communication architecture that connects remotely stationed telemedicine nodes and village clinics with remote doctors in specialty hospitals. Second, we present the development of disease severity pattern discovery and summarization algorithms, the result of which is a Consensus Abnormality Motif (CAM) and an associated Alert Measure Index, which suggests the immediacy of the patient data for doctor's consultative time. By frequently sending CAM as SMS in the absence of data network, we ensure timely delivery of critical data. Through a Detailed Data on Demand (DD-on-D) pull data mechanism doctors can further investigate complete data from the cloud. The CAM and DD-on-D mechanisms result in energy savings of up to 25%, while the data usage is reduced tremendously. Furthermore, we present a pilot deployment of the systems using a continuous cardiac monitoring device coupled with an intervention framework including more than 60 telemedicine nodes station in villages across India. © 2016 IEEE.