Study on a Low-Complexity ECG Compression Scheme With Two-Tier Sensors

Abstract

The industry of wearable remote health monitoring system keeps growing. In the diagnosis of cardiovascular disease, Electrocardiography (ECG) waveform is one of the major tools which is thus widely taken as the monitoring objective. For the purpose of reducing bit expenditure in the monitoring systems, we study the compression of ECG signal and propose a new compressor in low complexity. Different from the traditional ECG compressors, most of which are built on a single sensor, our compression scheme is based on multiple ECG sensors. The multi-sensor based compression scheme is able to provide more accurate sensing results. Besides the investigation into the structure of the compressor, we also jointly optimize the period and the bit number per sample in the transmission of ECG signal. Experiments are performed on records in MIT-BIH Arrhythmis database and European ST-T database. Experimental results show that our method outperforms conventional ones with respect to ECG reconstruction accuracy at the same bit rate consumption.

Multi-Distance Dispersion Entropy for ECG Signal Classification

“…The electrocardiogram (ECG) is one of the most frequently used techniques for diagnosing heart disease [1], [2]. The ECG records the heart's electrical activity over time by placing electrodes on the skin [3]. This signal is utilized to diagnose heart health and can be accessible by grasping the essentials of its distinctive waves: T, U, P, and QRS complex [4]- [6].…”