Electrocardiographic signal de-noising method based on adaptive threshold wavelet transform

An adaptive threshold, ECG signal technology, applied in applications, electrical digital data processing, special data processing applications, etc., can solve the problems of strangling wavelet coefficients, poor reconstruction and smoothing effect, easy to produce oscillations, etc. Separation, high practical value, the effect of retaining detailed information

CN108158573AActive Publication Date: 2018-06-15智慧康源(厦门)科技有限公司
10 Cites 25 Cited by

Patent Information

Authority / Receiving Office
CN · China
Current Assignee / Owner
Publication Date
2018-06-15

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention discloses an electrocardiographic signal de-noising method based on adaptive threshold wavelet transform. The method is characterized by comprising following steps: step 1: using the Mallat algorithm, the wavelet function sym6 and the number of decomposition layers J are selected, and the noisy ECG signal is decomposed by wavelet to obtain approximate coefficients and detail coefficients; step 2: setting the threshold for adaptive detail coefficients at each layer and selecting the threshold function; step 3: performing adaptive threshold processing on the detail coefficients ofeach layer, removing power frequency interference and myoelectric interference, and removing baseline drift by processing the approximation coefficients; step 4: performing wavelet reconstruction on the electrocardiographic signals after processing to obtain approximate optimal estimate value of signals. The method of the present invention makes full use of the multiresolution feature of the wavelet transform. An adaptive threshold selection method is provided. Different thresholds are used at each level to separate the noise and signal flexibly, improving the separability of signal characteristics; in the three aspects of visual, mean square error, and signal-to-noise ratio, the effect is better than the traditional method, and the detailed information of the image is retained better, which has higher practical value.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The invention relates to the field of electrocardiographic signal noise reduction, in particular to an electrocardiographic signal noise reduction method based on adaptive threshold wavelet transform. Background technique

[0002] Heart disease and various cardiovascular diseases are the most fatal and disabling diseases in the world. Their suddenness and unpredictability greatly restrict the diagnosis and treatment. With the aging population and the continuous improvement of living standards in our country, The incidence of heart disease and various cardiovascular diseases is on the rise. my country has a large population and relatively scarce doctors and medical equipment. Carrying out intelligent monitoring of ECG based on big data can effectively reduce the rate of death and disability. Noise reduction preprocessing of electrical signals is a key step; on the other hand, ECG signals are weak physiological signals with a frequency range of 0.5-150Hz, ...

Examples

Embodiment 1

[0040] The ECG signal denoising method based on adaptive threshold wavelet transform of the present invention, the operation in Matlab comprises:

[0041] Step 1: Obtain noisy ECG signals: select the first 1800 data of No. 118e_6 data in the MIT-BIH noise database to carry out qualitative analysis experiments; draw the original waveform diagram of the selected data in Matlab ( figure 2 ) and spectrogram ( image 3 ), as a comparative reference figure after using the traditional method and the method of the present invention for denoising respectively;

[0042] Step 2: select the wavelet function φ(x) and the number of decomposition layers J, and carry out wavelet decomposition to the noisy ECG signal: the present invention selects the sym6 wavelet function similar to the ECG signal form; 118e_6 data sampling rate 360Hz, useful signal 90 % is concentrated in 0.5-40Hz, and the baseline drift interference is mainly concentrated in the low frequency part within 1Hz, so 8-layer w...

Embodiment 2

[0065] ECG signal noise reduction method based on adaptive threshold wavelet transform, the operation in Matlab includes:

[0066] Step 1: Obtain the noisy ECG signal, select 1800 data from 20 seconds to 25 seconds of No. 103 data in the MIT-BIH arrhythmia database as the original "pure" signal, and superimpose three common interferences on it. First superimpose a sinusoidal signal with a frequency of 60Hz and an amplitude of 0.02mV to simulate power frequency interference, then use the awgn() function to superimpose random Gaussian white noise with a specified signal-to-noise ratio to simulate EMG, and finally superimpose the frequency at 0.3Hz and an amplitude of 0.2mV The sine signal simulates the baseline drift, so as to obtain the experimental data to be denoised; when superimposing random white noise, white noise with different signal-to-noise ratios is added, and the processing results are different. In this embodiment, the signal-to-noise ratio after superimposed noise ...