Real-Time ECG Noise Reduction with QRS Complex Detection for Mobile Health Services

Abstract

Cardiovascular disease is a serious threat to human life, especially when a sudden attack occurs, so real-time patient monitoring is crucial. Recent advances in health care and technology have led to equipment such as mobile micro-electro-mechanical systems, which can be used for more accessible public healthcare services. Electrocardiogram (ECG) data are traditionally used to investigate and monitor heart activities. However, the necessary electronic logic tags and (wireless) signal transmissions in a mobile healthcare device are susceptible to noise, which can result in false interpretations. Consequently, this study proposed a novel, low-complexity method for generating an optimized ECG wave suitable for mobile architecture. We first apply a bi-quad, high-pass filter to adjust baseline drifts. Then, a Savitzky–Golay filter smoothes the raw ECG, and moving variance and integral filters with thresholds are used to determine the QRS complex. We compared the results of the proposed technique to those from the moving average, Savitzky–Golay, PRASMMA, and Pan–Tompkins algorithms, using the well-known QT and MIT-BIH databases, and human subjects. The method was implemented on a mobile device integrating an open ECG platform as a prototype for real-time ECG monitoring systems.