Context-aware cardiac monitoring for early detection of heart diseases

The aim of this paper is to propose a scalable context-aware framework for early detection of several cardiovascular diseases by continuous monitoring using smart sensors and utilizing the strength of cloud computing. By constant sampling of ECG signal, vital signs, and activities our system detects possible symptoms of heart disease and alerts user by delivering context-aware service using flexible output modalities. A non-context-aware system that makes a decision based only on abnormal ECG signal can generate false alerts at high rate. Our proposed solution aims to reduce that rate by bringing different contexts in decision making process. As a proof of concept, we developed a simulated prototype to detect long term health risk of Premature Atrial Contraction (PAC), a common form of cardiac arrhythmia. The system can classify ECG signals as PAC using appropriate feature selection and learning algorithm. By tracking the stored context history and personal profile in the cloud database, our system detects smoking habit, alcohol consumption, caffeine intake of the user. It can also detect activities like stress, hypertension, and anxiety using different physiological parameters of the user and capable of sending situational warning notifications. Thus, this model can be a new mechanism for heart disease detection.