MODWT-based dynamic electrocardiogram real-time heart rate estimation method

Abstract

The invention discloses an MODWT-based dynamic electrocardiogram real-time heart rate estimation method. The method comprises the following steps: a) preprocessing; b)carrying out filtering processingon the dynamic electrocardiogram based on MODWT and each characteristic waveform frequency band characteristic, and (c) determining the main wave position and the real-time heart rate of the dynamicelectrocardiogram based on an electrocardio rule The method estimates the real-time heart rate of the dynamic electrocardiogram based on the MODWT and the electrocardio rule, and is easy to understandand implement and high in precision.

Description

technical field [0001] The invention proposes a MODWT-based real-time heart rate estimation method of a dynamic electrocardiogram, which relates to the field of electrocardiogram intelligent diagnosis. Background technique [0002] Electrocardiogram examination is a common item in physical examination. If the patient is suspected of having arrhythmia symptoms, he usually goes to the hospital to have an electrocardiogram. A wearable ECG detector, worn for a day, two weeks or longer, will generate dozens or hundreds of hours of ECG, the so-called Holter ECG, which will present two problems: first, in order to detect fatal For heart disease, the electrocardiogram detector should be able to analyze in real time and give important heart disease indicators; second: doctors need to check the long-term electrocardiogram every second, which is very time-consuming. For this reason, a better way of thinking is, Firstly, the electrocardiogram detector provides real-time important heart...

AI Technical Summary

Problems solved by technology

[0002] Electrocardiogram examination is a common item in physical examination. If the patient is suspected of having arrhythmia symptoms, he usually goes to the hospital to have an electrocardiogram. A wearable ECG detector, worn for a day, two weeks or longer, will generate dozens or hundreds of hours of ECG, the so-called Holter ECG, which will present two problems: first, in order to detect fatal For heart disease, the electrocardiogram detector should be able to analyze in real time and give important heart disease indicators; second: doctors need to check the long-term electrocardiogram every second, which is very time-consuming. For this reason, a better way of thinking is, Firstly, the electrocardiogram detector provides real-time important heart indicators for timely warning, and secondly, the electrocardiogram detector makes intelligent judgments to find out the time and location of possible arrhythmias, and then the doctor conducts further diagnosis on these places. Obviously, The primary task of real-time detection of dynamic electrocardiogram signals and auxiliary diagnosis results is to detect changes in heart rate, that is, to quickly and accurately identify the position of the main wave. The original intention of the invention of this patent comes from this

[0003] Compared with the conventional electrocardiogram, the ambulatory electrocardiogram is more susceptible to the interference of external factors. There are a lot of power frequency interference, baseline drift and electromyographic noise in the signal. At present, the low-pass filtering method is generally used to suppress the noise of the electrocardiogram signal. The shortcoming is that the fixed frequency band filtering cannot adapt to the randomness and non-stationarity of the ECG signal. The wavelet transform has good time-frequency local analysis characteristics and can solve these problems well. However, in the background of strong noise, the wavelet transform Finally, there is no translation invariance, and the Gibbs oscillation phenomenon often occurs at the Q and S waves with small amplitudes, resulting in signal distortion. Therefore, we propose a MODWT ECG signal adaptive filtering method. MODWT is a highly Redundant non-orthogonal wavelet transform has no requirement on sample size, has translation invariance of wavelet coefficients and scale coefficients, and no phase distortion, thereby retaining useful characteristic waveforms in ECG signals and avoiding the loss of useful ECG information

Images