ECG Signal Processing

Abstract

Electrocardiography (ECG) is a diagnostic method that allows sensing and record the electric activity of heart [. The measurement of electrical activity is used as a standard twelve-point system. At each of these leads to measure the useful signal and interference was measured. The intensity of interference depends on the artefacts (electrical lines, brum, motion artefacts, muscle, interference from the environment, etc.). For correct evaluation of measured signal there is a need to processing the measured signal to suitable form. At present, the use of electrocardiograms with sensors with contact scanning are difficult to set a time so we decided to use the principle of non-contact sensing. Such a device to measure the ECG was constructed under the project. The disadvantage of such devices is a problem with a high level of noise, which degrades a useful signal. The aim of this article is to pre-process the signals obtained from non-contact sensing. The contactless devices are powered from the network and battery. The electrodes were connected by way of Eithoven bipolar leads. Signals were pre-treated with suitable filters so that they are also appropriate for their subsequent analysis. In the filtration ECG signals was used as a method of linear (low pass filter, high pass, IIR (Infinite Impulse Response) peak, notch filter. The results of many signals clearly demonstrate removing noise in the ECG signals to the point that is also suitable for their analysis.

P- and T-wave delineation in ECG signals using parametric mixture Gaussian and dynamic programming

“…However, the ˇ prior helps in avoiding such erroneous R peaks since the ˇ PDF has very small value when n is close to zero or L. b. It is known that the heart rate and also the locations of P/T-peaks do not change drastically from one cycle to the next [2]. We assume that better estimates of the P/T-peaks could be obtained by incorporating this smoothness in the peak locations across cycles.…”

“…Most of the clinically useful information in ECGs can be obtained from the intervals, amplitudes and the shape of the ECG waveforms [1]. The development of accurate and robust methods for automatically extracting such waveform characteristics from ECG is a subject of major importance, especially for the analysis of long recordings [2].…”