CN111657916A - A method for analyzing atrial-ventricular synchrony signal - Google Patents

Abstract

A method for analyzing atrial and ventricular synchrony signals, the first step is signal acquisition, and an electrocardiogram signal of a subject is collected; the second step is signal processing, and the third step is to extract a P wave signal; and the fourth step, the calculation of the P wave correlation, using processing The good ECG signal calculates the P wave correlation of each p wave position, and adds a sliding window mechanism; step 5, detects the synchronization according to the threshold value, and judges the synchronization between the atrium and the ventricle according to the set threshold value; the operation of the present invention It is simple and can be used for real-time analysis of atrial-ventricular synchrony. It can analyze and detect atrial and ventricular asynchrony diseases such as atrial fibrillation, and obtain good detection results. Good method.

Description

A method for analyzing atrial-ventricular synchrony signal

technical field

The invention belongs to the technical field of biomedical signal processing, relates to an electrocardiographic signal processing and a signal correlation analysis method, and particularly relates to an atrium ventricular synchronization signal analysis method.

Background technique

The heart is one of the most important organs in the human body, and it is also very important for the study of the heart. In the electrophysiological activities of the heart, under normal circumstances, the sinoatrial node can spontaneously generate action potentials, and the sinoatrial node can generate an action potential at a rate of 60- per minute. The 100-beat rhythm generates electrical impulses, which first rapidly spread to the left and right atria, and then transmit the impulses to the Beaufort fibers of the ventricle through the atrioventricular node, which is the only pathway between the atrioventricular and ventricle. There is a delay in impulse conduction in the AV node, which takes about 0.15 seconds. The impulse then travels rapidly throughout the ventricle along the Purdenser fiber system, activating the entire ventricle in 0.1 seconds, causing the ventricular muscle to contract synchronously and expel blood. This appears in the ECG as a P wave followed by a QRS complex. The activity of the atria and ventricles shows a time-delayed synchrony, but in abnormal cardiac activity, such as atrial fibrillation, atrial cells produce frequencies as high as Abnormal electrical excitation signals of 400 to 600 times/min. This physiological phenomenon is manifested in the electrocardiogram as the disappearance of the P wave and the appearance of a series of high-frequency f waves of varying sizes, shapes and directions. About 350 to 600 times per minute. Atrial and ventricular synchrony. At present, most of the studies on AV synchrony are complex and require the assistance of multiple devices. Deng Yan et al. used real-time three-dimensional echocardiography to evaluate left atrial synchrony in dilated cardiomyopathy and achieved good results; Wang Dongmei et al also used three-dimensional echocardiography to evaluate left atrial synchrony in patients with heart failure and permanent atrial fibrillation. Synchronization, the results show that this method can perform atrial synchrony analysis very well, but its operation is complicated and cannot analyze synchrony in real time. At present, the diagnosis of atrial-ventricular asynchrony type diseases such as atrial fibrillation based on ECG is mostly performed by doctors. The artificial discrimination cannot give the result in real time. Therefore, we need an algorithm that can analyze the atrioventricular synchronization in real time. The algorithm can automatically determine the synchronization of the atrioventricular signal in real time, and can be applied to life scenarios such as families. Behavior is detected and may eventually give early warning signs.

SUMMARY OF THE INVENTION

In order to overcome the above-mentioned defects of the prior art, the purpose of the present invention is to provide a P-wave correlation-based signal analysis method for atrial-ventricular synchrony, which is simple in calculation, accurate and efficient, and can be used for real-time detection of the synchronization of atrial and interventricular signals It is of great significance for the analysis and detection of diseases such as clinical atrial fibrillation and other cardiac synchrony disorders, and can be applied to various scenarios such as families, which has a broad application value.

In order to achieve the above object, the technical scheme adopted in the present invention is:

An atria-ventricular synchronization signal analysis method, comprising the following steps:

Step 1: Signal Acquisition

Use an electrocardiogram (ECG) acquisition device to collect continuous ECG data from a single lead of a patient;

Step 2: Signal Processing

The collected ECG data are preprocessed to obtain a noise-free original ECG signal, and then the position of the QRS wave peak is detected to obtain the time point of each R wave peak position;

Step 3: P wave extraction

For the position of the R wave crest obtained in step 2, determine the position of the beginning and end of the P wave;

The extraction of the P wave in step 3 is as follows: for the first R wave peak position extracted in step 2, move to the left for 160ms to obtain the end position of the P wave, and then continue to move to the left for 110ms to obtain the start position of the P wave. This obtains the P wave signal, which is recorded as the P ₁ signal, and thus obtains the following P wave signals P ₂ , P ₃ . . . P _i , where i represents the i-th P wave signal;

Step 4: Calculation of P-Wave Correlation

的相关性C_i：For the i-th P-wave signal, calculate the mean value of the P-wave signal and the previous i P-wave signals

The correlation C _i of :

的计算公式如下:When calculating the correlation coefficient, where

The calculation formula is as follows:

计算相关系数,然后再取这些结果中最大的相关系数作为最终的该处P波相关性值，其中每次滑动的距离win为

fs表示信号的采样频率；而设距离P_i左侧1个窗宽win位置处的信号为P_i(-1)，距离P_i右侧1个窗宽win位置处的信号为P₍₁₎，分别取P_i左侧共l个信号即P_i(-1),P_i(-2),,,P_i(-l)，取P_i右侧共r个信号即P_i(1),P_i(2),,,P_i(r)，其中l或r取值应该在3到15之间，然后分别求这l+1+r个信号与

的相关系数C_i(k)，其中-l≤k≤r,且k为整数,对于某一段信号P_i(k),该信号由序列(x₁,x₂…x_n)构成，而对于

信号，其构成序列为(y₁,y₂…y_n),那么相关系数C_i(k)的计算公式如下：Take the signals at multiple positions around P _i respectively and

Calculate the correlation coefficient, and then take the largest correlation coefficient among these results as the final P wave correlation value at this place, where the distance win of each sliding is

fs represents the sampling frequency of the signal; and the signal at the position of 1 window width win to the left of Pi is P _{i (-1)} , and the signal at the position of 1 window width win to the right of _Pi is P ₍₁₎ , respectively take a total of l signals on the left side of Pi, namely Pi _(-1) , Pi _(-2) ,,,,P _{i (-l)} , and take a total of r signals on the right side of _Pi , namely Pi ₍₁₎ ,P _i(2) ,,,P _i(r) , where the value of l or r should be between 3 and 15, and then find these l+1+r signals and

The correlation coefficient C _i(k) of , where -l≤k≤r, and k is an integer, for a certain segment of signal P _i(k) , the signal consists of a sequence (x ₁ ,x ₂ …x _n ), and for a certain segment of signal P i(k)

signal, its composition sequence is (y ₁ , y ₂ ... y _n ), then the calculation formula of the correlation coefficient C _i(k) is as follows:

Take the maximum value of all the results as the final correlation coefficient, denoted as C _i , as the P wave characteristic value of the heartbeat at this point;

Step 5: Detect Synchronization Based on Threshold

For each P wave correlation C _i obtained in step 4, set a correlation threshold M, if C _i < M, mark this segment of the signal as an abnormal segment, where the value of M is between 0.6 and 0.7. , which can effectively distinguish normal P wave signals from abnormal signals; dynamically calculate the ratio d of abnormal segments to the total number of segments in the one minute before the end of this segment of P waves, and set a threshold N, the value range of N is 0.3 Between 0.5, if d>N, it indicates that the signal synchronization between the atrium and the ventricle is poor, otherwise it is normal.

The present invention overcomes the shortcomings of the existing atrioventricular synchrony analysis methods, and proposes a new, easy-to-implement, simple calculation, and real-time analysis method for atrioventricular synchrony. The invention applies the P wave signal in the ECG signal, can calculate the correlation between each P wave signal and the mean signal in real time, and analyzes the synchronization of the atrial and ventricular signals in a certain period of time according to the value of the correlation, and the result is accurate. Well, it is of great significance for the analysis and detection of diseases of clinical atrial fibrillation and other cardiac synchrony disorders.

Description of drawings

Fig. 1 is the overall structure diagram of the algorithm of the present invention.

Figure 2 is a schematic diagram of the calculation of the P-wave correlation.

FIG. 3 is a schematic diagram of detecting synchronization according to a threshold.

specific embodiment

In order to illustrate the operation process of the present invention more clearly, the present invention will be described in detail below with reference to the accompanying drawings and examples.

Referring to Fig. 1, a method for analyzing an atria-ventricular synchrony signal includes the following steps:

Step 1: Signal acquisition:

Continuous ECG data from a single lead of a patient is acquired using an electrocardiogram (ECG) acquisition device.

Step 2: Signal Processing

A zero-phase-shift high-pass filter is used to filter out the DC component in the original ECG signal, and then the peak position of the QRS wave is detected to obtain the time point of each R wave peak position.

Step 3: P wave extraction

For the first R wave peak position extracted in step 2, move to the left for 160ms to obtain the end position of the P wave, and then continue to move to the left for 110ms to obtain the start position of the P wave, thereby obtaining the P wave signal, which is recorded as P ₁ signal, and thus obtain the following P wave signals P ₂ , P ₃ . . . P _i , where i represents the i-th P wave signal.

Step 4: P wave correlation calculation

的相关性，其中

的计算公式如下:Referring to Fig. 2, for the P wave signal obtained in step 3, calculate the average value of this P wave signal and the first i P wave signals

correlation, where

The calculation formula is as follows:

的相关系数时，加入一个位置滑动机制。该种机制的基本思想是：在计算相关系数时，并不只计算P_i这一个位置上的信号和

之间的相关系数，而是分别取P_i左右的多个位置处的信号分别与

计算相关系数,然后再取这些结果中最大的相关系数作为最终的该处P波相关性值，其中每次滑动的距离win为

其中fs表示心电信号的采样率，设距离P_i左侧1个窗宽win位置处的信号为P_i(-1)，距离P_i右侧1个窗宽win位置处的信号为P₍₁₎，分别取P_i左侧共l个信号即P_i(-1),P_i(-2),,,P_i(-l)，取P_i右侧共r个信号即P_i(1),P_i(2),,,P_i(r)，本方法中，取l＝6，r＝8。然后分别求这l+1+r个信号与

的相关系数C_i(k)，其中-l≤k≤r,且k为整数,对于某一段信号P_i(k),该信号由序列(x₁,x₂…x_n)构成，而对于

信号，其构成序列为(y₁,y₂…y_n),那么相关系数C_i(k)的计算公式如下：Since the method used to determine the position of the P wave corresponding to each heartbeat can only roughly estimate the position of the P wave, there is a certain deviation from the actual position of the P wave, and such deviation may further lead to inaccurate correlation coefficients between the calculated signals. . Therefore, in actual operation, in order to reduce this deviation and increase the fault tolerance rate, when calculating _Pi and

When the correlation coefficient is , a position sliding mechanism is added. The basic idea of this mechanism is: when calculating the correlation coefficient, not only the signal sum at the position P _i is calculated

The correlation coefficient between , but take the signals at multiple positions around P _i respectively and

Calculate the correlation coefficient, and then take the largest correlation coefficient among these results as the final P wave correlation value at this place, where the distance win of each sliding is

Where fs represents the sampling rate of the ECG signal, and the signal at the position of 1 window width win to the left of Pi is P _{i (-1)} , and the signal at the position of 1 window width win to the right of _Pi is P _{( 1)} , respectively take a total of l signals on the left side of _Pi , namely Pi _(-1) , Pi _(-2) ,,, _{Pi (-l)} , and take a total of r signals on the right side of Pi, namely Pi ₍ -1) ₁₎ ,P _i(2) ,,,P _i(r) , in this method, take l=6, r=8. Then find these l+1+r signals and

The correlation coefficient C _i(k) of , where -l≤k≤r, and k is an integer, for a certain segment of signal P _i(k) , the signal consists of a sequence (x ₁ ,x ₂ …x _n ), and for a certain segment of signal P i(k)

signal, its composition sequence is (y ₁ , y ₂ ... y _n ), then the calculation formula of the correlation coefficient C _i(k) is as follows:

Take the maximum value of all the results as the final correlation coefficient, denoted as C _i , as the eigenvalue of the P wave of the heartbeat at this point.

Step 5: Detect Synchronization Based on Threshold

3, for each P wave correlation C _i obtained in step 4, set a correlation threshold M=0.7, if C _i <M, then mark this segment of signal as an abnormal segment, otherwise it means that the When the P wave signal is normal, the ratio d of the abnormal segment to the total number of segments per minute is dynamically calculated, and a threshold value N=0.4 is set. If d>N, it means that the synchronization between the atrium and the ventricle is poor in this time period; otherwise, it means normal. .

Claims (3)

1. An atrial ventricular synchronous signal analysis method is characterized by comprising the following steps:

the method comprises the following steps: signal acquisition

Acquiring continuous electrocardiographic data of a single lead of a patient by an Electrocardiographic (ECG) acquisition device;

step two: signal processing

Preprocessing the acquired electrocardio data to obtain an original electrocardio signal without noise, and then detecting the position of a QRS wave peak to obtain the time point of each R wave peak position;

step three: p-wave extraction

Determining the positions of the wave crest of the R wave obtained in the step two, and determining the starting and ending positions of the P wave;

step four: calculation of P-wave correlation

For the ith P-wave signal, calculating the mean value of the P-wave signal and the first i P-wave signals

Correlation of (2)_i；

Step five: detecting synchronicity according to a threshold

For each P wave correlation C obtained in step four_iSetting a correlation threshold M, if C_iIf M is less than M, the segment signal is marked as an abnormal segment, where M has a value of 0.6 toWhen the frequency is between 0.7, normal P wave signals and abnormal signals can be effectively distinguished; dynamically calculating the ratio d of the abnormal segment to the total segment number in the previous minute taking the segment P wave as the end, and setting a threshold value N, wherein the value range of N is between 0.3 and 0.5, if d is larger than N, the signal synchronism of the atria and the ventricles is poor, otherwise, the signal synchronism is normal.

2. The ANS-ventricular synchronous signal analysis method of claim 1,

the P wave extraction in the third step is specifically as follows: for the first R wave peak position extracted in the step two, moving the position to the left for 160ms to obtain the end position of the P wave, then continuing moving the position to the left for 110ms to obtain the start position of the P wave, thereby obtaining a P wave signal which is recorded as P₁The signal, and thus the following P-wave signal P₂、P₃……P_iWhere i represents the ith P-wave signal.

3. The ANS-ventricular synchronous signal analysis method of claim 1,

the fourth step is specifically as follows: in the calculation of the correlation coefficient, wherein

The calculation formula of (a) is as follows:

get P separately_iSignals at a plurality of positions on the left and right are respectively compared with

Calculating correlation coefficients, and taking the maximum correlation coefficient in the results as the final P wave correlation value, wherein the distance win of each sliding is

fs meterIndicating the sampling frequency of the signal; and set a distance P_iThe signal at the position of 1 window width win on the left side is P_i(-1)Distance P_iThe signal at the position of 1 window width win on the right side is P₍₁₎Respectively take P_iLeft side total of one signal, P_i(-1)，P_i(-2)，，，P_i(-l)Taking P_iR signals on the right, P_i(1)，P_i(2)，，，P_i(r)Wherein l or r should be between 3 and 15, and then respectively summing the l +1+ r signals

Coefficient of correlation C_i(k)Where-l ≦ k ≦ r, and k is an integer, for a certain segment of the signal P_i(k)The signal consisting of a sequence (x)₁，x₂...x_n) Is formed as a whole

A signal having a composition sequence of (y)₁，y₂...y_n) Then the correlation coefficient C_i(k)The calculation formula of (a) is as follows:

taking the maximum value in all the results as the final correlation coefficient and marking as C_iThe characteristic value of the P wave of the heart beat is obtained.

Images