Using casual reasoning for anomaly detection among ECG live data streams in ubiquitous healthcare monitoring systems

Anomalies in cardiac functionality can be fatal. Early detection of these anomalies, and in many cases their precursors, can save lives. The probability of the occurrence of these anomalies is extremely among people with a pre-diagnosed heart condition. In this research, we discovered that much remote Electrocardiography (ECG) monitoring systems do not convey “enough” information to the diagnosing doctor or the nominated caregiver. A few examples of this information can be the type of cardiac abnormality, the exact waveform of the ECG signal, time and frequency of the occurrence of the anomaly, machine-understandable part so that medical SCADA be alerted about the case, and immediate preventative urgent steps correlated to that emergency. It is also important to delivery surrounding context to the Health Information System so that the medical expert make his/her diagnosis with ample support data. The most important component in this communication is the security of contents from cybercrimes. We propose a cost-efficient and non-invasive health monitoring system that is secure and quickly deployable. The presented system embeds an intelligent wearable data acquisition system with unique identification algorithms requiring very little computational time and simple threshold-based classification.