Remote ECG Monitoring Kit to Predict Patient-Specific Heart Abnormalities

摘要

Electrocardiogram (ECG) signals are widely used to check heart rhythms and general health conditions using low-cost and relatively accurate equipment. However, the majority of commercial off-the-shelf ECG kits are generic and their normal ranges are set based on an average normal hearth signal hence ignorant of extreme variations among different people's normal heart signals. As such, many false alarms are generated if the global thresholds are selected too strict. On the other hand, loosely selected thresholds may result in missing many true alarms. Furthermore, the kits output report typically includes a limited number of basic parameters such as heart rate and hence negligent to a rich set of information exploitable from signal morphology. In this paper, we developed a prototype for patient-specific heart monitoring kit, which learns the properties of a patient's normal ECG signal over time and reports significant deviations from this normal behavior. In order to reduce the false alarm rate even further, the kit is equipped with an accelerometer in order to distinguish between high heart rates due to high physical activity levels and real abnormalities. This personalized remote heat monitoring kit with the proposed signal processing and self-tuning capabilities provides more detailed information and insightful interpretations compared to generic devices, therefore can be used for remote health monitoring of high-risk and elderly people.