CN113633293B - Heart-derived sudden death early warning method for chaotically detecting T-wave electricity alternation - Google Patents

Abstract

The invention discloses a method for early warning of sudden cardiac death for chaotic detection of T-wave electrical alternation. The detection method calculates the correlation dimension and the maximum Lyapunov exponent; calculates the first error and the second error, the first error is the absolute value of the difference between the correlation dimension and the average value of the correlation dimension, and the second error is the largest The absolute value of the difference between the Lyapunov exponent and the average value of the maximum Lyapunov exponent; judge whether the first error is less than the first threshold, and judge whether the second error is less than the second threshold; when the first error is less than the first threshold, or the second If the error is less than the second threshold, a sudden cardiac death warning is issued. The invention can search for chaotic features with high feasibility for TWA detection, effectively detect TWA, and provide better early warning for sudden cardiac death.

Description

Sudden cardiac death early warning method based on chaos detection of T wave electrical alternation

technical field

The invention relates to the technical field of biomedicine, in particular to a method for early warning of sudden cardiac death for chaotic detection of T wave electrical alternation.

Background technique

Sudden Cardiac Death (SCD) is defined as loss of consciousness within a maximum of one hour after the onset of symptoms without any potentially fatal preconditions, the main clinical manifestation is sudden loss of consciousness, rapid, fatal, non-traumatic Physiological death, and the out-of-hospital treatment rate is extremely low. In recent years, there have been cases of famous people who died of sudden cardiac death due to ineffective rescue, and the current sudden cardiac death is no longer exclusive to the elderly, gradually showing signs of younger age, and many young people born in the 1990s also Inevitably, this is obviously a wake-up call to mankind. Sudden cardiac death is a sudden onset, and once it occurs, patients often lose their lives without timely rescue. At this time, if the signs of sudden cardiac death can be captured in advance, it will be possible to obtain sufficient time for the patient’s follow-up treatment, avoid the situation where the rescue is not timely due to sudden sudden death, and effectively improve the survival rate of the patient. Therefore, , the early detection of the risk of sudden cardiac death has become an urgent need.

Sudden cardiac death is a sudden onset, and once it occurs, patients often lose their lives without timely rescue. At this time, if the signs of sudden cardiac death can be captured in advance, we can gain sufficient time for the follow-up treatment of the patient, avoid the situation where the rescue is not timely due to sudden death, and effectively improve the survival rate of the patient. According to the relevant literature, T Wave Alternans (TWA) is a pathological cardiac activity. As a tool for predicting sudden cardiac death, TWA has high research value and broad application prospects.

SUMMARY OF THE INVENTION

The present invention aims to at least solve the technical problems existing in the prior art. Therefore, the present invention proposes a sudden cardiac death early warning method for chaotic detection of T-wave electrical alternation, which can search for chaotic features with high feasibility for TWA detection, effectively detect TWA, and achieve a better response to sudden cardiac death. early warning.

The present invention also provides a sudden cardiac death early warning system for chaotic detection of T wave electrical alternation with the above-mentioned chaotic detection T wave electrical alternation early warning method for sudden cardiac death.

The present invention also provides a computer-readable storage medium.

In a first aspect, this embodiment provides a method for early warning of sudden cardiac death for chaotic detection of T-wave electrical alternation, including the following steps:

Obtain ECG data;

performing noise removal on the ECG data;

Using the chaos detection method to calculate the T-wave alternation data of the ECG data to obtain the correlation dimension and the maximum Lyapunov exponent;

Calculate the first error and the second error, the first error is the absolute value of the difference between the correlation dimension and the average value of the correlation dimension, and the second error is the maximum Lyapunov exponent the absolute value of the difference from the mean of the maximum Lyapunov exponent;

judging whether the first error is less than a first threshold, and judging whether the second error is less than a second threshold;

When the first error is smaller than the first threshold, or the second error is smaller than the second threshold, a sudden cardiac death warning is issued.

The method for early warning of sudden cardiac death for chaotic detection of T-wave electrical alternation according to the embodiment of the present invention has at least the following beneficial effects:

First obtain ECG data, and then preprocess the ECG data to remove interference noise, mainly including EMG interference, power frequency interference and baseline drift. EMG signal is an almost unavoidable interference signal, which is caused by human activities, The tension and tremor of muscles in other parts are mainly caused by the interference caused by muscle twitches in the area where the electrode pads are located when collecting ECG data. The irregularity is a kind of Gaussian white noise, which belongs to high-frequency interference, and the frequency is generally distributed in the range of 30Hz~ Between 2000Hz, the ECG waveform is a kind of ripple with fast change rate, small and irregular. The power frequency interference is caused by the power supply connected to the ECG acquisition equipment and the external electromagnetic field. It belongs to the AC signal. It is mainly reflected in the superposition of obvious sine waves on the ECG. The concrete point is that there will be many tiny burrs on the ECG waveform. Baseline drift is generally caused by factors such as human respiration, slight displacement of electrocardiograph electrode pads, and skin surface impedance. Baseline drift will cause the signal data segment to float or distort along with the baseline, the change is relatively slow, and the frequency is low, generally less than 1Hz. This kind of interference has a great impact on the follow-up ECG analysis research, especially on the accuracy of ST segment identification. Since T wave alternation appears in the ST band on the ECG waveform, baseline drift must be dealt with as much as possible to reduce the impact on TWA detection.

Then use the chaos detection method to calculate the correlation dimension and the maximum Lyapunov exponent for the T wave electrical alternation data of the ECG data; calculate the first error and the second error, and the first error is the sum of the correlation dimension and the average value of the correlation dimension The absolute value of the difference between the two, the second error is the absolute value of the difference between the maximum Lyapunov index and the average value of the maximum Lyapunov index; judge whether the first error is less than the first threshold, and judge whether the second error is less than the second threshold ; When the first error is smaller than the first threshold, or the second error is smaller than the second threshold, a sudden cardiac death warning is issued. The method for early warning of sudden cardiac death for chaotic detection of T-wave electrical alternation provided in this embodiment can search for chaotic features with high feasibility for TWA detection, effectively detect TWA, and achieve better early warning of sudden cardiac death.

According to some embodiments of the present invention, before the preprocessing of the ECG data to remove interference noise, the method further includes the steps of:

Select the lead and sampling time point of the ECG data;

Visualize the ECG data.

According to some embodiments of the present invention, the preprocessing of the ECG data to remove interference noise includes steps;

The ECG data are preprocessed to remove EMG interference, power frequency interference and to correct baseline drift.

According to some embodiments of the present invention, the preprocessing of the ECG data to remove myoelectric interference, power frequency interference and baseline drift includes steps;

filtering out the EMG interference using a Butterworth low-pass filter;

Use a 50Hz or 60Hz power frequency notch filter to filter out the power frequency interference;

The baseline drift was corrected using the median filtered Kaiser window method.

According to some embodiments of the present invention, before the correlation dimension and the maximum Lyapunov exponent are obtained by calculating the T wave electrical alternation data of the ECG data using the chaos detection method, the method further includes the steps:

Select the power spectrum to perform phase-space reconstruction of the time series.

According to some embodiments of the present invention, after judging whether the first error is smaller than a first threshold and judging whether the second error is smaller than a second threshold, the steps include:

When the first error is greater than the first threshold and the second error is greater than the second threshold, the process ends.

In a second aspect, this embodiment provides a sudden cardiac death early warning system for chaotic detection of T-wave electrical alternation, including: a memory, a processor, and a computer program stored in the memory and running on the processor, the processing When the computer executes the computer program, the method for early warning of sudden cardiac death for detecting T wave electrical alternation in chaos according to the first aspect is realized.

According to the embodiment of the present invention, the early warning system for sudden cardiac death for detecting T-wave electrical alternation in chaos has at least the following beneficial effects: The T-wave alternation early warning method for sudden cardiac death can find chaotic features with high feasibility for TWA detection, effectively detect TWA, and achieve better early warning of sudden cardiac death.

In a third aspect, this embodiment provides a computer-readable storage medium, where the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to cause a computer to execute the chaos described in the first aspect A method for early warning of sudden cardiac death by detecting T wave alternation.

Additional aspects and advantages of the present invention will be set forth, in part, from the following description, and in part will be apparent from the following description, or may be learned by practice of the invention.

Description of drawings

The above and/or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, the abstract of which is intended to be in full conformity with one of the accompanying drawings of the specification:

1 is a flowchart of a method for early warning of sudden cardiac death for chaotic detection of T-wave electrical alternation provided by an embodiment of the present invention;

2 is a TWA electrocardiogram of a method for early warning of sudden cardiac death for chaotic detection of T-wave electrical alternation provided by another embodiment of the present invention;

Fig. 3 is the S-T wave result diagram after filtering and removing P wave of the method for early warning of sudden cardiac death for chaotic detection of T wave electrical alternation provided by another embodiment of the present invention;

Fig. 4 is the S-T wave result figure after filtering and removing P wave of the method for early warning of sudden cardiac death for chaotic detection of T wave electrical alternation provided by another embodiment of the present invention;

5 is a diagram showing the result of S-T wave after filtering and removing P wave in a method for early warning of sudden cardiac death for chaotic detection of T wave electrical alternation provided by another embodiment of the present invention;

6 is a case diagram of phase space reconstruction of a method for early warning of sudden cardiac death for chaotic detection of T-wave electrical alternation provided by another embodiment of the present invention;

Fig. 7 is the TWA electrocardiographic data solving correlation dimension result diagram of the method for early warning of sudden cardiac death for chaotic detection of T wave electrical alternation provided by another embodiment of the present invention;

FIG. 8 is a structural diagram of a sudden cardiac death early warning system for chaotic detection of T-wave electrical alternation provided by another embodiment of the present invention.

Detailed ways

In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, but not to limit the present invention.

It should be noted that although the functional modules are divided in the schematic diagram of the system and the logical sequence is shown in the flowchart, in some cases, the modules can be divided differently from the system, or executed in the order in the flowchart. steps shown or described. The terms "first", "second" and the like in the description and claims and the above drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

In recent years, TWA has been regarded as a predictor of SCD, and it has also confirmed that TWA as a non-invasive detection method has great promise in predicting sudden cardiac death. However, TWA is an abnormal phenomenon that is difficult to observe with the naked eye in the human body surface electrocardiogram, and is easily disturbed or even annihilated by various kinds of noise. Therefore, accurate detection of TWA in ECG is a challenging problem. According to the detection principle, TWA detection methods are classified into time domain detection methods and transform domain detection methods.

The time domain detection method mainly analyzes the signal qualitatively and quantitatively through the symbol transformation in the time domain, and compares the TWA time domain diagnostic criteria to judge the existence of TWA. Correlation analysis is common. The correlation analysis method takes the ECG signal in the continuous cardiac cycle as the object, calculates the average amplitude of all T waves in the signal, and uses the information such as the cross-correlation and auto-correlation function between the current amplitude and the average amplitude of the T wave to construct. The correlation index, which is used as the detection index of TWA, is recorded as the alternating correlation index.

The TWA transform domain detection method is based on short-time Fourier transform or spectral transform. The Spectral Method (SM) mainly divides the digitized ECG signal into segments according to each heartbeat, and regards them as a vector, and then performs spectrum analysis and TWA detection on the vector signal in a certain window.

Non-linear principle detection TWA method: Laplace likelihood ratio method, the idea is to establish a non-Gaussian TWA probability model, and analyze TWA through the model, but the shortcomings of this method are obvious poor anti-noise and low robustness.

The invention provides a sudden cardiac death early warning method for chaotic detection of T wave electrical alternation, which can search for chaotic features with high feasibility for TWA detection, effectively detect TWA, and achieve better early warning of sudden cardiac death. .

The embodiments of the present invention will be further described below with reference to the accompanying drawings.

Referring to FIG. 1 , FIG. 1 is a flowchart of a method for early warning of sudden cardiac death for chaotic detection of T wave electrical alternation provided by an embodiment of the present invention, and the method for early warning of sudden cardiac death for chaotic detection of T wave electrical alternation includes but is not limited to step S110 Go to step S160.

Step S110, acquiring ECG data;

Step S120, performing noise removal on the ECG data;

Step S130, using the chaos detection method to calculate the T-wave alternation data of the ECG data to obtain the correlation dimension and the maximum Lyapunov exponent;

Step S140, calculate and obtain the first error and the second error, the first error is the absolute value of the difference between the correlation dimension and the average value of the correlation dimension, and the second error is the average value of the maximum Lyapunov exponent and the maximum Lyapunov exponent the absolute value of the difference between;

Step S150, judging whether the first error is smaller than the first threshold, and judging whether the second error is smaller than the second threshold;

Step S160, when the first error is smaller than the first threshold, or the second error is smaller than the second threshold, a sudden cardiac death warning is issued.

In one embodiment, ECG data is obtained first, and then the ECG data is preprocessed to remove interference noise, which mainly includes EMG interference, power frequency interference and baseline drift. EMG signals are almost unavoidable interference signals. , which is caused by human activity and the tension and vibration of muscles in other parts, mainly due to the interference caused by the muscle twitching in the area where the electrode pads are located when collecting ECG data. The frequency is generally distributed between 30Hz and 2000Hz. What is reflected in the ECG waveform is a ripple with a fast change rate and small irregularities. The power frequency interference is caused by the power supply connected to the ECG acquisition equipment and the external electromagnetic field. It belongs to the AC signal. It is mainly reflected in the superposition of obvious sine waves on the ECG. The concrete point is that there will be many tiny burrs on the ECG waveform. Baseline drift is generally caused by factors such as human respiration, slight displacement of electrocardiograph electrode pads, and skin surface impedance. Baseline drift will cause the signal data segment to float or distort along with the baseline, the change is relatively slow, and the frequency is low, generally less than 1Hz. This kind of interference has a great impact on the follow-up ECG analysis research, especially on the accuracy of ST segment identification. Since T wave alternation appears in the ST band on the ECG waveform, baseline drift must be dealt with as much as possible to reduce the impact on TWA detection. Then use the chaos detection method to calculate the correlation dimension and the maximum Lyapunov exponent for the T wave electrical alternation data of the ECG data; calculate the first error and the second error, and the first error is the sum of the correlation dimension and the average value of the correlation dimension The absolute value of the difference between the two, the second error is the absolute value of the difference between the maximum Lyapunov index and the average value of the maximum Lyapunov index; judge whether the first error is less than the first threshold, and judge whether the second error is less than the second threshold ; When the first error is smaller than the first threshold, or the second error is smaller than the second threshold, a sudden cardiac death warning is issued.

In one embodiment, ECG data is acquired, leads and sampling time points of the ECG data are selected, the ECG data is visualized, the ECG data is preprocessed to remove noise, and Using the chaos detection method to calculate the correlation dimension and the maximum Lyapunov exponent, the first error and the second error are calculated, and the first error is the difference between the correlation dimension and the average value of the correlation dimension. Absolute value, the second error is the absolute value of the difference between the maximum Lyapunov exponent and the average value of the maximum Lyapunov exponent; judge whether the first error is less than the first threshold, and judge whether the second error is less than the second threshold; when the first error If it is less than the first threshold, or the second error is less than the second threshold, a sudden cardiac death warning is issued.

Referring to FIG. 2 to FIG. 5, FIG. 2 is a TWA electrocardiogram of a method for early warning of sudden cardiac death with chaotic detection of T-wave electrical alternation provided by another embodiment of the present invention, and FIG. 3 is a T-wave detection of chaos provided by another embodiment of the present invention. Figure 4 shows the result of filtering and removing the P wave in the electrical alternating sudden cardiac death early warning method after filtering and removing the P wave. Figure 5 is a result diagram of the S-T wave after filtering and removing the P wave in the method for early warning of sudden cardiac death for chaotic detection of T wave electrical alternation provided by another embodiment of the present invention.

The ECG data is taken from the T-band. T-wave is an important component of the ECG, which reflects the repolarization process of the ventricular myocardium. T-wave alternation is a kind of abnormal T-wave. Studies have shown that there is a direct connection between sudden cardiac death and T-wave alternation. The possibility of sudden cardiac death can be predicted in advance by detecting T-wave alternation. In the ECG data, compared with the R wave, the characteristics of the T wave are not obvious and the changes are subtle. The effective detection of the weak signal by the chaotic system can just solve the problem of the abnormal T wave.

To detect the electrical alternation of T wave, it is necessary to intercept the S-T wave band first. This paper will use the wavelet packet algorithm to determine the time domain boundary of the T wave of the ECG by extracting the waveform of the frequency range of the T wave of the ECG, and extract the T wave in the single-cycle ECG.

According to the analysis of the spectrum of the QRS complex and T wave of the electrocardiogram, the bandwidth frequency of the QRS wave is concentrated at 0 to 38 Hz, accumulating nearly 99% of the energy, and the peak energy of the QRS wave is concentrated in the vicinity of 8 to 16 Hz; the bandwidth of the T wave is 0 ～8Hz, the peak energy is concentrated in the frequency range of 1～8Hz.

The wavelet packet algorithm is a generalization of wavelet decomposition, which can effectively decompose the signal into signals of different frequencies according to the frequency, and the re-superposition and combination of these signals will become the original signal. According to the frequency distribution of the complex bandwidth of the ECG, the S-T band is mainly distributed in the frequency range of 0 to 8 Hz. Through wavelet packet decomposition, for a sample signal with a sampling frequency of mHz, the total signal S is expressed as:

The decomposed signal frequency bands are:

Taking the EuropeanST-TDatabase dataset as an example, its 10-second data length is 2500, and its frequency is 250HZ. If it is decomposed into 5 layers, the root node (5,0) frequency band is 0-8HZ, which is the S-T band.

In one embodiment, the electrocardiographic data is acquired; the electrocardiographic data is preprocessed to remove interference noise; the T wave electrical alternation data of the electrocardiographic data is calculated by the chaos detection method to obtain the correlation dimension and the maximum Lyapunov exponent; Error and second error, the first error is the absolute value of the difference between the associated dimension and the mean of the associated dimension, and the second error is the absolute value of the difference between the maximum Lyapunov exponent and the mean of the maximum Lyapunov exponent value; determine whether the first error is smaller than the first threshold, and determine whether the second error is smaller than the second threshold; when the first error is smaller than the first threshold, or the second error is smaller than the second threshold, a sudden cardiac death warning is issued. Among them, preprocessing ECG data to remove interference noise includes removing EMG interference, removing power frequency interference and removing baseline drift. It is mainly caused by the muscle twitching in the area where the electrode pads are located when collecting ECG data. The irregularity is a kind of Gaussian white noise, which belongs to high-frequency interference. The frequency is generally distributed between 30Hz and 2000Hz. On the electrical waveform is a ripple with a fast rate of change, small and irregular. The power frequency interference is caused by the power supply connected to the ECG acquisition equipment and the external electromagnetic field. It belongs to the AC signal. It is mainly reflected in the superposition of obvious sine waves on the ECG. The concrete point is that there will be many tiny burrs on the ECG waveform. The power frequency in some areas is generally 50Hz, while the power frequency in the United States is 60Hz.

Baseline drift is generally caused by factors such as human respiration, slight displacement of electrocardiograph electrode pads, and skin surface impedance. 1Hz, this kind of interference has a great impact on the follow-up ECG analysis research, especially on the accuracy of ST segment identification. Since the T wave alternation appears in the ST band on the ECG waveform, the baseline drifts It must be disposed of as much as possible to reduce the impact on TWA detection.

It can be understood that the Butterworth low-pass filter is used to remove EMG interference. The principle of the low-pass filter is that the frequency response will decrease with the increase of frequency, which can maximize the internal frequency of the passband. As the signal flows, the stopband will approach 0 more and more as the frequency increases. The power frequency in most parts of China is generally 50Hz, while the power frequency in the United States is 60Hz. To remove the power frequency interference, use a 50/60Hz power frequency notch filter. The notch filter has distinct characteristics and low design difficulty. The calculation speed of the interference effect is also fast. Taking the data of the international standard database as an example, the frequency is 60Hz according to the American standard, and the design stopband cutoff frequencies are 59Hz and 61Hz. The problem of baseline drift can be effectively corrected by using the median filter Kaiser window function method.

Assuming that the ECG waveform signal is x(n), 0≤n≤N-1, use the median filter with a window width of L=2k+1 to process x(n), and extend the k points at the beginning and end of the signal waveform:

For the new continuation signal x(n) obtained by processing, -k≤n≤N+k-1, perform median filtering point by point from n=0 to n=N-1 to obtain the baseline drift:

X(n)=med[x(n-k),x(n-k+1),...,x(n),x(n+1),...,x(n+k)],n=0,1 ,…,N-1 where X(n) is the baseline drift waveform, and med[·] is the median operator. Subtract X(n) from the original waveform x(n) to get the signal after filtering out the baseline drift:

Referring to FIG. 6, FIG. 6 is a case diagram of a phase space reconstruction of a method for early warning of sudden cardiac death with chaotic detection of T-wave electrical alternation provided by another embodiment of the present invention; FIG. 7 is a chaotic detection T provided by another embodiment of the present invention. Correlation dimension result diagram of TWA ECG data solution of wave-electric alternating sudden cardiac death early warning method.

In an embodiment, before using the chaos detection method to calculate the correlation dimension and the maximum Lyapunov exponent for the T wave electrical alternation data of the ECG data, the method further includes the steps of: selecting a power spectrum and reconstructing the time series in phase space.

Phase space reconstruction: In order to expand the hidden chaotic attractor as much as possible, because the chaotic attractor can often reflect the chaotic regularity that is difficult to capture in the original time series. Chaos theory believes that each component in the system is related to each other, and the information of the relevant component is implicit in the development process of any component. In order to obtain complete and accurate qualitative information of the system, it is often necessary to understand comprehensive and sufficient state evolution information.

Assuming a given time series {x(i), i=1,2,3,...,N}, the sequence length is N, based on two important parameters (delay time τ and embedding dimension m), for {x( i)} perform phase space reconstruction, and the new phase space matrix is a matrix composed of high-dimensional vectors:

The phase points of N _m phase spaces can be expressed as:

X(i)=[x(i),x(i+τ),x(i+2τ),…,x(i+(m-1)τ)]i=1,2,…,M

To reconstruct a suitable phase space, the key lies in the selection of delay time and embedding dimension m:

(1) Regarding the delay time τ: If τ is too small, the correlation between the delay variables may be too close, so that the time series is only changed from a one-dimensional space to a multi-dimensional sequence group of the same variables, and the hidden feature information behind it cannot be explored. . On the contrary, if the selection is too large, the situation of reconstruction will show a more severe self-folding phenomenon.

(2) About the embedding dimension m: if m is too large, the amount of calculation will increase significantly, which will greatly increase the unnecessary workload and reduce the efficiency of the work. If m is too low, the existing shortcomings will be highlighted, the attractor will self-intersect, the opening is incomplete, and the dynamic behavior of the system cannot be accurately and fully characterized.

The CC algorithm is used in this experiment. The idea is that when the time series contains noise and the length is limited, the two are interrelated parameters, which are inseparable and need to be determined at the same time. The algorithm is simple in operation with limited samples, small in computation, greatly reduces computation time, and has strong anti-noise ability. The delay time τ depends on the time window τ _w =(m-1)τ, and τ and τ _w are estimated at the same time by correlation integration, and then the embedding dimension m is determined.

The original time series is x(i), i=1,2,...,N, which is divided into t disjoint subsequences, and the defined length is:

l=[N/t]

In the formula, [·] means rounding. The t subsequences are expanded as follows:

Calculate the statistic S(m, N, r, τ) for each subsequence separately:

In the formula, C _l represents the correlation integral of the l-th subsequence, which can be used to characterize that the radius of the neighborhood is larger than that in the phase space.

The probability of the distance between any two points is defined as:

In the formula, M=N-(m-1)t, r represents the spatial distance threshold, and θ(x) is Heaviside

Step function, the distance between space vectors X _i and X _j adopts infinite norm.

When N→∞, there are:

According to BDS statistics, if the time series is independent and identically distributed, when N→∞, S(m,r,τ) is always 0. S(m, r, τ) ~ τ reflects the autocorrelation of the sequence, when its first zero-crossing point or the minimum difference for all r values,

The trajectory of the reconstructed chaotic attractor in the phase space is completely opened.

Define the maximum deviation ΔS(m,τ) with respect to r:

ΔS(m,τ)=max{S(m,r _i ,τ)}-min{S(m,r _i ,τ)}

In the formula, ΔS(m, τ) is used to measure the maximum deviation of S(m, r, τ) ~ τ for all r.

The optimal delay time takes the first zero point of S(m,r,τ)～τ or the first minimum point of ΔS(m,τ)～τ.

第一个零点或

的第一个局部极小点，定位这两点的时间，一turn up

the first zero or

The first local minimum point of , the time to locate these two points, a

Generally, these two times are taken as the optimal delay time τ. The embedded window width τ _w is the time corresponding to the global minimum point selected from S _cor (t) as τ _w . Further, the optimal embedding dimension m can be determined.

m=τ _w /τ+1

In one embodiment, the correlation dimension and the maximum Lyapunov exponent are calculated to obtain the correlation dimension and the maximum Lyapunov exponent for the T wave electrical alternation data of the ECG data, including the steps:

The first space vector and the second space vector are selected, the correlation integral is calculated according to the first space vector and the second space vector, and the correlation dimension is calculated according to the correlation integral.

In one embodiment, after judging whether the first error is smaller than the first threshold and judging whether the second error is smaller than the second threshold, the steps include:

When the first error is greater than the first threshold and the second error is greater than the second threshold, the process ends.

It is understandable that the calculation of the correlation dimension: the attractor trajectories in the phase space of the chaotic system will form a self-similar structure with infinite nesting, and this phenomenon can be described by the correlation dimension. Therefore, the correlation dimension is regarded as an important parameter of chaos detection and analysis.

After reconstructing the phase space of any time series {x(i), i=1, 2, 3, ..., N}, select any two space vectors X(i) and X(j) to count the difference between the two The number of point pairs whose distance is less than or equal to any given neighborhood radius r accounts for the proportion of all point pairs, that is, the associated integral:

In the formula, M=N-(m-1)t, r represents the spatial distance threshold, θ(x) is the Heaviside step function, and the distance between the spatial vectors X _i and X _j adopts the infinite norm. For a sufficiently small r correlation integral satisfying C(r)∝r ^D , the growth trend is exponential, and the correlation dimension is defined:

D can be estimated from the slope of the linear part of the curve lnC(r) ~ ln(r):

As the dimension of the phase space continues to increase, but the value of D does not change along with it, it can be regarded as the associated dimension D. This method is simple and easy to implement, and it is widely used, but it is sensitive to noise, and sometimes there is a large error in the calculation results.

Calculation of the maximum Lyapunov exponent: The Lyapunov exponent refers to the average divergence rate index of adjacent trajectories in the phase space. It can be known from chaos theory that as long as the Lyapunov exponent has a positive value, it indicates the generation of chaos, and the larger the value, the stronger the degree of chaos. On the contrary, if the Lyapunov exponent is negative, it can indicate that the system is in an ordered state, and the motion is in a stable and convergent state. Therefore, it is possible to check whether the system is chaotic by the positive and negative conditions of the Lyapunov exponent. After getting a reconstructed phase space Y(t _i ):

Y(t _i )=[x(t _i ),x(t _i+τ ),…,x(t _i+(m-1)τ )]i=1,2,…,N

Take the initial point Y(t ₀ ) and find its nearest neighbor, denoted as Y ₀ (t ₀ ). Let its initial distance from the nearest adjacent point Y ₀ (t ₀ ) be L ₀ , that is, the minimum distance between two points. Trace the time evolution of these two phase points until time t ₁ ,

The distance between two phase points exceeds a certain value, which is defined as follows:

L' ₀ =|Y(t ₁ )-Y ₀ (t ₁ )|>ε

When ε>0, keep Y(t ₁ ), and find another phase point Y ₁ (t ₁ ) near Y(t ₁ ), so that

L ₁ =|Y(t ₁ )-Y ₁ (t ₁ )|<ε

Continue to repeat the above process until the point of traversing the entire sequence, Y(t) reaches the end point N of the time series.

Assuming that the tracking evolution process is iterated M times in total, the maximum Lyapunov exponent (LLE) is:

The delay time τ and the embedding dimension m are calculated by the C-C algorithm to reconstruct the phase space. Some of the results are as follows:

The data show that the cardiac system is in a chaotic state, and the morbid state is significantly weaker than the normal chaotic state. Another important information is also obtained, that is, the minimum of TWA patients is significantly different from the other three categories, so the maximum Lyapunov exponent is selected together with the correlation dimension as the chaotic feature of TWA detection.

The design of sudden cardiac death detection based on chaotic features has a significant degree of discrimination in two chaotic features (correlation dimension and maximum Lyapunov exponent) for TWA patients, healthy people and patients with other diseases.

Define the error value ε ₁ :

代表着关联维数的平均值，D′代表着与

差值最大的关联维数值。in,

Represents the average value of the associated dimension, and D' represents the

The associated dimension value with the largest difference.

Likewise, define the error value ε ₂ :

代表着最大Lyapunov指数的平均值，L′代表着与

差值最大的最大Lyapunov指数。in,

represents the mean value of the largest Lyapunov exponent, and L′ represents the

The largest Lyapunov exponent with the largest difference.

为0.877，得关联维数的误差值

同理，最大Lyapunov指数

均值为0.0423，其误差值

将ε₁、ε₂作为TWA的检测指标。Calculate the correlation dimension D of TWA patients in the existing international database (100 groups are selected in this paper) and calculate the mean value to obtain the correlation dimension of TWA patients

is 0.877, the error value of the associated dimension is obtained

Similarly, the maximum Lyapunov exponent

The mean is 0.0423, and its error value

ε ₁ and ε ₂ are used as detection indicators of TWA.

与待测时间序列关联维数D₁之间差的绝对值和最大Lyapunov指数平均值

与待测时间序列最大Lyapunov指数L₁之间差的绝对值。若满足

和

即可判定该待测心电数据属于TWA数据而做出警示，从而达到心源性猝死的预警目的。Assuming that a piece of ECG data to be measured {x(i), i=1,2,3,...,N} is input, the correlation dimension and the maximum Lyapunov exponent value are calculated through chaos detection, and the correlation dimension D ₁ of the time series is obtained. and the largest Lyapunov exponent L ₁ , calculate the mean of the associated dimension

The absolute value of the difference between the dimension D ₁ associated with the time series under test and the mean value of the maximum Lyapunov exponent

The absolute value of the difference with the maximum Lyapunov exponent L ₁ of the time series to be tested. if satisfied

and

It can be determined that the ECG data to be measured belongs to TWA data and an alert is made, so as to achieve the purpose of early warning of sudden cardiac death.

Referring to FIG. 8 , FIG. 8 is a structural diagram of a sudden cardiac death early warning system for chaotic detection of T wave electrical alternation provided by another embodiment of the present invention.

The present invention also provides a sudden cardiac death early warning system for chaotic detection of T wave electrical alternation, comprising: a memory, a processor and a computer program stored in the memory and running on the processor, the processor executing the The computer program realizes the above-mentioned sudden cardiac death early warning method of chaotic detection of T wave electrical alternation. The early warning system for sudden cardiac death with chaotic detection of T wave electrical alternation applies the sudden cardiac death early warning method for chaotic detection of T wave electrical alternation as described in the first aspect, which can find chaotic features with high feasibility for TWA detection. TWA can perform effective detection and better early warning of sudden cardiac death.

In addition, an embodiment of the present invention also provides a computer-readable storage medium storing computer-executable instructions, the computer-executable instructions being executed by one or more control processors, eg, controlling The processor can execute the method steps S110 to S160 in FIG. 1 .

Those of ordinary skill in the art can understand that all or some of the steps and systems in the methods disclosed above can be implemented as software, firmware, hardware, and appropriate combinations thereof. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit . Such software may be distributed on computer-readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). As is known to those of ordinary skill in the art, the term computer storage media includes both volatile and nonvolatile implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data flexible, removable and non-removable media. Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cartridges, magnetic tape, magnetic disk storage or other magnetic storage devices, or may Any other medium used to store desired information and which can be accessed by a computer. In addition, communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and can include any information delivery media, as is well known to those of ordinary skill in the art .

The preferred implementation of the present invention has been specifically described above, but the present invention is not limited to the above-mentioned embodiments. Those skilled in the art can also make various equivalent deformations or replacements on the premise of not violating the spirit of the present invention. These Equivalent modifications or substitutions are included within the scope defined by the claims of the present invention.

Claims (8)

1. a chaotic detection T-wave electrical alternation of a sudden cardiac death early warning system, comprising: memory, a processor and a computer program that is stored on the memory and can be run on the processor, it is characterized in that, the processor executes all When the computer program is described, the following steps are implemented: Obtain ECG data from human body surface ECG; performing noise removal on the ECG data; Using the chaos detection method to calculate the T wave electrical alternation data of the ECG data to obtain the correlation dimension D ₁ of the ECG data and the maximum Lyapunov exponent L ₁ of the ECG data; Obtain the association dimension D and the maximum Lyapunov exponent L of the historical TWA patients, and calculate the average value of the association dimension based on the association dimension D

Calculate the mean of the largest Lyapunov exponent based on the largest Lyapunov exponent L

Average based on correlation dimension

and the mean of the largest Lyapunov exponent

The first error ε ₁ and the second error ε ₂ are obtained by calculation; wherein:

Among them, D' represents the average of the association dimension D and the association dimension of the historical TWA patients

The relationship dimension with the largest difference between

The largest Lyapunov exponent with the largest difference between; |.| represents the absolute value function; calculate

and

when

and

Then, it is determined that the ECG data belongs to TWA, and a sudden cardiac death warning is issued. 2. The sudden cardiac death early warning system of chaotic detection T wave electrical alternation according to claim 1, is characterized in that, before the described electrocardiographic data is preprocessed to remove interference noise, it also comprises the step: Select the lead and sampling time point of the ECG data; Visualize the ECG data. 3. The sudden cardiac death early warning system for chaotic detection of T-wave electrical alternation according to claim 1, characterized in that, the described electrocardiographic data is preprocessed to remove interference noise, comprising steps; The ECG data are preprocessed to remove EMG interference, power frequency interference and to correct baseline drift. 4. The early warning system of sudden cardiac death according to claim 3, characterized in that, described electrocardiographic data is preprocessed to remove electromyographic interference, power frequency interference and baseline drift, includes steps; filtering out the EMG interference using a Butterworth low-pass filter; Use a 50Hz or 60Hz power frequency notch filter to filter out the power frequency interference; The baseline drift was corrected using the median filtered Kaiser window method. 5. The early warning system of sudden cardiac death for chaotic detection of T-wave electrical alternation according to claim 1, characterized in that, in the T-wave electrical alternation data of the ECG data, the chaotic detection method is used to calculate and obtain the described Before the correlation dimension D ₁ of the electrocardiographic data and the maximum Lyapunov exponent L ₁ of the electrocardiographic data, further steps are included: Select the power spectrum to perform phase-space reconstruction of the time series. 6 . The early warning system of sudden cardiac death for chaotic detection of T-wave electrical alternation according to claim 5 , wherein, the T-wave electrical alternation data of the ECG data is calculated to obtain the cardiac arrest by using a chaotic detection method. Correlation dimension D ₁ of electrical data, including steps: Select the first space vector and the second space vector from the phase space reconstruction, calculate the correlation integral according to the first space vector and the second space vector, and calculate the correlation of the ECG data according to the correlation integral dimension D ₁ . 7. The early warning system for sudden cardiac death with chaotic detection of T wave electrical alternation according to claim 1, characterized in that, in the calculation

and

After that, also include steps: when

and

end process. 8. A computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to cause a computer to perform the following steps: Obtain ECG data from human body surface ECG; performing noise removal on the ECG data; Using the chaos detection method to calculate the T wave electrical alternation data of the ECG data to obtain the correlation dimension D ₁ of the ECG data and the maximum Lyapunov exponent L ₁ of the ECG data; Obtain the association dimension D and the maximum Lyapunov exponent L of the historical TWA patients, and calculate the average value of the association dimension based on the association dimension D

Calculate the mean of the largest Lyapunov exponent based on the largest Lyapunov exponent L

Average based on correlation dimension

and the mean of the largest Lyapunov exponent

The first error ε ₁ and the second error ε ₂ are obtained by calculation; wherein:

Among them, D' represents the average of the association dimension D and the association dimension of the historical TWA patients

The relationship dimension with the largest difference between

The largest Lyapunov exponent with the largest difference between; |.| represents the absolute value function; calculate

and

when

and

Then, it is determined that the ECG data belongs to TWA, and a sudden cardiac death warning is issued.

Images