A CPR Artifact Suppression Method, System, Electronic Device, and Storage Medium

Abstract

The present invention discloses a method, system, electronic device and storage medium for suppressing CPR artifacts, including the following steps: S1: Obtain the original ECG data containing CPR artifacts to be processed; S2: Preprocess the original ECG data to separate the ECG signal and the CPR artifacts for the first time, and obtain the CPR artifact signal without the ECG signal; S3: Calculate the power spectral density of the CPR artifact signal in S2; S4: Find out the component frequency points of the CPR artifact signal according to the power spectral density PSD, and then reconstruct the CPR artifact signal according to the component frequency points to obtain the CPR interference signal; S5: Use a Kalman adaptive filter to recursively solve the CPR interference signal to obtain the CPR artifact estimation; S6: Remove the CPR artifact estimation from the original ECG data to obtain the ECG data without CPR artifacts. Through this method, CPR artifacts can be removed from the original ECG data signal, and the accuracy of defibrillation judgment types can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electrocardiogram signal processing, and particularly relates to a CPR artifact suppression method, system, electronic device and storage medium. Background Art

[0002] When the human body is severely impacted externally or the condition of its own diseases changes, the heart and lungs may stop beating, which will cause damage to the human body. In modern first aid techniques, cardiopulmonary resuscitation (CPR) is generally performed on patients by pressing, and at the same time, a defibrillator analyzes and displays the electrocardiogram signal of the patient for the staff to visually check and decide whether to perform defibrillation.

[0003] However, during continuous external chest compressions, CPR (cardiopulmonary resuscitation) artifact signals will cause great interference to the original electrocardiogram (ECG) signal, which will affect the analysis of the electrocardiogram rhythm by the defibrillator, reduce the accuracy of the defibrillation analysis result, and may lead to the inability to correctly distinguish between defibrillable / non-defibrillable categories. Summary of the Invention

[0004] Aiming at the technical problem of CPR artifact interfering with electrocardiogram signals in the prior art, the present invention proposes a CPR artifact suppression method, system, electronic device and storage medium. The original electrocardiogram signal is initially separated by a band-pass filter, then the frequency points of the main frequency components are used to reconstruct the phase CPR interference signal, and then the Kalman adaptive filter is used for recursive estimation to obtain the reconstructed CPR interference signal, and finally the CPR interference signal is removed from the original electrocardiogram signal.

[0005] In order to achieve the above object, the present invention provides the following technical solutions:

[0006] A CPR artifact suppression method, comprising the following steps:

[0007] S1: Obtain the original ECG data containing CPR artifacts to be processed;

[0008] S2: Preprocess the original ECG data to perform the first separation of the ECG signal and the CPR artifact, and obtain the CPR artifact signal without the ECG signal;

[0009] S3: Calculate the power spectral density of the CPR artifact signal in S2;

[0010] S4: Find out the component frequency points of the CPR artifact signal according to the power spectral density PSD, and then reconstruct the CPR artifact signal according to the component frequency points to obtain the CPR interference signal;

[0011] S5: Use the Kalman adaptive filter to recursively solve the CPR interference signal to obtain the CPR artifact estimator;

[0012] S6: Remove the CPR artifact estimator from the original ECG data to obtain the ECG data without CPR artifacts.

[0013] Preferably, in S1, the type of the original ECG data includes at least two waveforms; each waveform includes at least 100 samples, and each sample has at least 1000 sampling points.

[0014] Preferably, in S2, the preprocessing method is as follows:

[0015] Use a Butterworth first-order filter to design a band-pass filter with a band-pass frequency of [1 Hz, 3 Hz], and perform low-pass filtering on the original ECG data to remove the ECG signal therein, realizing the first separation of the ECG signal and the CPR artifact.

[0016] Preferably, S3 includes:

[0017] Step S31: Perform autocorrelation calculation on the CPR artifact signal:

[0018]

[0019] In formula (1), R x (τ) represents the autocorrelation result of performing autocorrelation calculation on x(t); t represents time; τ represents time delay; x(t) represents the time-domain signal at time t after passing through the band-pass filter; x(t + τ) represents the time-domain signal at time t + τ after passing through the band-pass filter;

[0020] Step S32: Perform Fourier transform on the autocorrelation result R x (τ):

[0021]

[0022] In formula (2), S x (f) represents the Fourier transform result; j represents the imaginary part of a complex number; f represents the signal sampling frequency; τ represents time delay;

[0023] Step S33: Integrate the Fourier transform result S x (f) to obtain the power spectral density:

[0024]

[0025] In formula (3), PSD represents the power spectral density; S x (f) represents the Fourier transform result.

[0026] Preferably, S4 includes:

[0027] S41: Calculate the frequency point of the maximum component of the CPR artifact signal according to the power spectral density:

[0028] f0 = getMaxPos(PSD) (4)

[0029] In formula (4), f0 represents the frequency point of the maximum component of the CPR artifact signal, that is, the first main frequency point; getMax represents the function to return the maximum value; getMaxPos represents the subscript of the maximum value of the power spectral density and the frequency point when the power spectral density is the maximum.

[0030] S42: Reconstruct the CPR artifact signal according to the component frequency point to obtain the CPR interference signal:

[0031] S cpr [n] = {A0[n]cos(ω0n + θ0) + K * A1[n] * cos(2ω0n + θ1)};

[0032] ω0 = 2πf0 (5)

[0033] In formula (5), S cpr [n] represents the reconstructed CPR interference signal; A0[n] represents the amplitude coefficient of the sine function constructed by the first main frequency point; ω0 represents the angular velocity; n represents the total amount of the signal (for example, if the signal is collected for 4 seconds and the sampling frequency is 250Hz, then the total amount of the signal is 1000); θ0 represents the initial phase of the sine function constructed by the first main frequency point; A1[n] represents the amplitude coefficient of the sine function constructed by the second main frequency point; θ1 represents the initial phase of the sine function constructed by the second main frequency point; f0 represents the frequency point of the maximum component of the CPR artifact signal.

[0034] K represents the coefficient to control the second main frequency point:

[0035] K = 1; if f0 < f c

[0036] K = 0; if f0 ≥ f c (6)

[0037] In formula (6), f c represents the cut-off frequency.

[0038] Preferably, the S5 includes:

[0039] S51: Establish a Kalman adaptive filter, and its model function is:

[0040] x[n + 1] = φ[n + 1, n] * x[n] + v1[n]

[0041] y[n] = C[n] * x[n] + v2[n] (7)

[0042] In formula (7), x[n + 1] represents the state parameter at time n + 1; φ[n + 1, n] represents the Kalman state transition matrix; x[n] represents the state parameter at time n; v1[n] represents white noise with a mean of 0;

[0043] y[n] represents the observed quantity, that is, the original ECG signal; C[n] represents the state transition matrix from x[n] to y[n]; v2[n] represents white noise with a mean of 0;

[0044] x[n] = (A0[n]cosθ0[n], A0[n]sinθ0[n], A1[n]cosθ1[n], A1[n]sinθ1[n]) T ;

[0045] C[n] = (cos(ω0n), -sin(ω0n), Kcos(2ω0n), -Ksin(2ω0n)) (8)

[0046] In formula (8), x[n] represents the state parameter at time n; A0[n] represents the amplitude coefficient of the sine function constructed by the first main frequency point; θ0[n] represents the initial phase of the sine function constructed by the first main frequency point; A1[n] represents the amplitude coefficient of the sine function constructed by the second main frequency point; θ1[n] represents the initial phase of the sine function constructed by the second main frequency point; T represents the transpose matrix; ω0 represents the angular velocity;

[0047] Step S52: By recursively solving the model function of the Kalman adaptive filter, the CPR artifact estimator S cpr [n]:

[0048] S cpr [n] = C[n] * x[n].

[0049] Preferably, in the above S6, the method for obtaining the ECG data without CPR artifacts is:

[0050] origin ECG = y[n] - S cpr [n] (9)

[0051] In formula (9), origin ECG represents the ECG data without CPR artifacts; y[n] represents the observed quantity, that is, the original ECG data; S cpr [n] represents the CPR artifact estimator.

[0052] The present invention also provides a CPR artifact suppression system, including:

[0053] An acquisition unit for acquiring the original ECG data containing CPR artifacts to be processed;

[0054] A preprocessing unit for preprocessing the original ECG data to obtain a CPR artifact signal without ECG signals;

[0055] A power spectral density calculation unit for calculating the power spectral density corresponding to the CPR artifact signal;

[0056] A phase reconstruction unit for calculating the component frequency points of the CPR artifact signal according to the power spectral density, and then reconstructing the CPR artifact signal according to the component frequency points to obtain a CPR interference signal;

[0057] An adaptive filter construction unit for constructing a Kalman adaptive filter to recursively solve the CPR interference signal to obtain a CPR artifact estimate;

[0058] An output unit for removing the CPR artifact from the original ECG data according to the CPR artifact estimate and outputting the ECG data without CPR artifacts.

[0059] The present invention also provides an electronic device, the electronic device includes a processor, and the processor is used to run a computer program stored in a memory so that the electronic device implements the steps of the above method.

[0060] The present invention also provides a computer-readable storage medium, in which a computer program is stored, and the computer program implements the steps of the above method when running on a processor.

[0061] In summary, due to the adoption of the above technical solution, compared with the prior art, the present invention has at least the following beneficial effects:

[0062] The present invention makes full use of the characteristic that the main frequency component of the ECG signal interfered by the CPR artifact is approximately the CPR artifact signal in the low-frequency band, and uses a band-pass filter to preliminarily separate and remove most of the ECG signals, retaining the CPR artifact signal; then uses the frequency points of the main frequency component to reconstruct the phase CPR interference signal, and then uses a Kalman adaptive filter for recursive estimation to obtain the CPR interference signal, and finally removes the CPR artifact from the original ECG data signal.

[0063] This method is completely completed by software, without pretending to be any hardware device, and does not rely on an external reference signal, which greatly improves the rhythm classification of the electrocardiogram signal with CPR artifact interference, thereby improving the accuracy of the defibrillation judgment type; and has a small amount of calculation and better CPR artifact suppression performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 It is a schematic diagram of a CPR artifact suppression method according to an exemplary embodiment of the present invention.

[0065] Figure 2 Schematic diagram of a CPR artifact suppression system according to an exemplary embodiment of the present invention. Detailed implementation manners

[0066] The present invention will be further described in detail below in conjunction with embodiments and specific implementation manners. However, this should not be construed as limiting the scope of the above-mentioned subject matter of the present invention to the following embodiments. Any technology implemented based on the content of the present invention belongs to the scope of the present invention.

[0067] In the description of the present invention, it should be understood that the orientation or positional relationships indicated by the terms "longitudinal", "lateral", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. are based on the orientation or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as limiting the present invention.

[0068] As Figure 1 shown, the present invention provides a CPR artifact suppression method, which specifically includes the following steps:

[0069] S1: Obtain the original ECG (electrocardiogram) data containing CPR (cardiopulmonary resuscitation) artifacts to be processed.

[0070] In this embodiment, the type of the original ECG data includes at least two waveforms, such as waveforms N and AF (which are the most main waveform types in the MIT and CU international authoritative electrocardiogram databases and are also the waveforms that need to be classified for defibrillation as stipulated in the national standard AED defibrillation classification standard); each waveform includes at least 100 samples, and each sample has at least 1000 sampling points.

[0071] In this embodiment, the sampling frequency of the original ECG data is 250HZ.

[0072] S2: Preprocess the original ECG data, that is, separate the ECG signal and the CPR artifact for the first time to obtain a CPR artifact signal that does not contain the ECG signal.

[0073] In this embodiment, since the original ECG data includes both the ECG signal and the CPR artifact, the defibrillator's judgment of whether to defibrillate based on the ECG signal will be interfered by the CPR artifact. Therefore, it is necessary to separate the CPR artifact from the original ECG data first.

[0074] In this embodiment, a Butterworth first-order filter is used to design a band-pass filter with a band-pass frequency of [1 Hz, 3 Hz] to perform low-pass filtering on the original ECG data to remove the ECG signal therein, realizing the first separation of the ECG signal and the CPR artifact.

[0075] According to statistical laws, the CPR energy frequency range is around 1 to 3 Hz, while the frequency of the ECG signal is generally greater than 3 Hz. Therefore, the ECG signal therein can be removed by a band-pass filter with a band-pass frequency of [1 Hz, 3 Hz], and the remaining data is the CPR artifact signal.

[0076] In this embodiment, when preprocessing the original ECG data, only the main energy is separated, and most of the ECG signals can be removed. However, there are still a small number of impurities in the obtained CPR artifact signal. Therefore, the CPR artifact signal needs to be further processed.

[0077] S3: Calculate the power spectral density of the CPR artifact signal in S2.

[0078] Step S31: Perform autocorrelation calculation on the CPR artifact signal:

[0079]

[0080] In formula (1), R x (τ) represents the autocorrelation result of performing autocorrelation calculation on x(t); t represents time; τ represents time delay, which needs to be determined during simulation to set what value has the best effect and belongs to a hyperparameter; x(t) represents the time-domain signal at time t after passing through the band-pass filter; x(t + τ) represents the time-domain signal at time t + τ after passing through the band-pass filter;

[0081] Step S32: Perform Fourier transform on the autocorrelation result R x (τ):

[0082]

[0083] In formula (2), S x (f) represents the Fourier transform result; j represents the imaginary part of a complex number; f represents the signal sampling frequency, for example, f = 250 Hz; τ represents the time delay.

[0084] Step S33: Integrate the Fourier transform result S x (f) to obtain the power spectral density:

[0085]

[0086] In formula (3), PSD represents the power spectral density; S x (f) represents the Fourier transform result.

[0087] In this embodiment, after obtaining the power spectral density of the CPR artifact signal, the CPR artifact function is represented as follows: the vertical axis is power density in units of dB / Hz, and the horizontal axis is frequency in units of Hz. The frequency point with the largest component of the power spectral density is the first dominant frequency point, and the frequency point with the second largest power spectral density is the second dominant frequency point. Generally speaking, most of the energy of the CPR signal is concentrated at the first dominant frequency point.

[0088] S4: Find the component frequency points of the CPR artifact signal based on the power spectral density PSD output in S3, and reconstruct the CPR artifact signal according to the component frequency points to obtain the CPR interference signal.

[0089] S41: There is a power spectral density PSD at each frequency point of the CPR artifact signal. If there is no energy, the corresponding PSD at that frequency point is 0. That is, the power spectral density PSD is not a single value but a series of values corresponding to different frequency points. Therefore, the maximum component frequency point of the CPR artifact signal can be calculated based on the power spectral density:

[0090] f0 = getMaxPos(PSD) (4)

[0091] In formula (4), f0 represents the maximum component frequency point of the CPR artifact signal, that is, the first dominant frequency point; getMax represents the function to return the maximum value; getMaxPos represents the subscript of the maximum value of the power spectral density and the frequency point when the power spectral density is the maximum.

[0092] S42: Reconstruct the CPR artifact signal according to the component frequency points to obtain the CPR interference signal.

[0093] S cpr [n] = {A0[n]cos(ω0n + θ0) + K * A1[n] * cos(2ω0n + θ1)};

[0094] ω0 = 2πf0 (5)

[0095] In formula (5), S cpr [n] represents the reconstructed CPR interference signal; A0[n] represents the amplitude coefficient of the sine function constructed at the first dominant frequency point; ω0 represents the angular velocity; n represents the total amount of the signal (for example, if the signal is collected for 4 seconds and the sampling frequency is 250 Hz, then the total amount of the signal is 1000); θ0 represents the initial phase of the sine function constructed at the first dominant frequency point; A1[n] represents the amplitude coefficient of the sine function constructed at the second dominant frequency point; θ1 represents the initial phase of the sine function constructed at the second dominant frequency point; f0 represents the maximum component frequency point of the CPR artifact signal;

[0096] K represents the coefficient controlling the second dominant frequency point:

[0097] K = 1; if f0 < fc

[0098] K = 0; if f0 ≥ f c (6)

[0099] In formula (6), f c represents the cut-off frequency, and can take a value of 1.6725 Hz.

[0100] In this embodiment, generally speaking, most of the energy of the CPR signal is concentrated on the first main frequency point. Therefore, generally only the physical quantity of the first main frequency point is considered for reconstruction. However, when the condition reaches f0 < f in formula (6) c the energy of the second main frequency point cannot be ignored. Therefore, the second main frequency point is also required for reconstruction. K = 1 indicates that the switch for reconstructing the second main frequency component is turned on.

[0101] S5: Use a Kalman adaptive filter to recursively solve the CPR interference signal to obtain the CPR artifact estimator.

[0102] S51: Establish a Kalman adaptive filter, and its model function is:

[0103] x[n + 1] = φ[n + 1, n] * x[n] + v1[n]

[0104] y[n] = C[n] * x[n] + v2[n] (7)

[0105] In formula (7), x[n + 1] represents the state parameter at time n + 1; φ[n + 1, n] represents the Kalman state transition matrix; x[n] represents the state parameter at time n; v1[n] represents white noise with a mean of 0;

[0106] y[n] represents the observed quantity, that is, the ECG signal interfered by the CPR artifact; C[n] represents the state transition matrix from x[n] to y[n]; v2[n] represents white noise with a mean of 0;

[0107] The meaning of formula (7) is to show that the state at time n + 1 is transferred from the state at the previous time n.

[0108] In this embodiment, y[n] is composed of two parts of signals:

[0109] y[n] = S ECG [n] + S CPR [n] (8)

[0110] In formula (8), S ECG [n] represents the ECG signal of the ECG without being interfered by the CPR artifact; S CPR [n] represents the ECG signal of the ECG interfered by the CPR artifact.

[0111] x[n] = (A0[n]cosθ0[n], A0[n]sinθ0[n], A1[n]cosθ1[n], A1[n]sinθ1[n]) T ;

[0112] C[n] = (cos(ω0n), -sin(ω0n), Kcos(2ω0n), -Ksin(2ω0n)) (9)

[0113] In formula (9), x[n] represents the state parameter at time n; A0[n] represents the amplitude coefficient of the sine function constructed by the first main frequency point; θ0[n] represents the initial phase of the sine function constructed by the first main frequency point; A1[n] represents the amplitude coefficient of the sine function constructed by the second main frequency point; θ1[n] represents the initial phase of the sine function constructed by the second main frequency point; T represents the transpose matrix; ω0 represents the angular velocity.

[0114] Step S52: By recursively solving the model function of the Kalman adaptive filter, the instantaneous coefficient of each sampling point is obtained, that is, recursively solved from n -> n + 1, and then the CPR artifact estimator S cpr [n]:

[0115] S cpr [n] = C[n] * x[n].[[]END]]

[0116] S6: According to the CPR artifact estimator S cpr [n], the original ECG data is processed to remove the CPR artifact, realizing the suppression of the CPR artifact; then the ECG data after suppressing the CPR artifact is used to judge the defibrillation / non - defibrillation category, thereby improving the accuracy of the defibrillation judgment.

[0117] origin ECG = y[n] - S cpr [n] (10)

[0118] In formula (10), origin ECG represents the ECG data without CPR artifact; y[n] represents the observed quantity, that is, the original ECG data in S1; S cpr [n] represents the CPR artifact estimator.

[0119] The present invention makes full use of the characteristic that the main frequency component of the ECG signal interfered by CPR artifacts is approximately the CPR artifact signal in the low-frequency band. A band-pass filter is used for preliminary separation to remove most of the ECG signals and retain the CPR artifact signals. Then, the frequency points of the main frequency component are used to reconstruct the phase CPR interference signal, and the Kalman adaptive filter is used for recursive estimation to obtain the CPR interference signal. Finally, the CPR artifacts are removed from the original ECG data. This method is completely completed by software, without pretending to use any hardware devices and not relying on external reference signals, greatly improving the rhythm classification of the electrocardiogram signals with CPR artifact interference, thereby improving the accuracy of defibrillation judgment types. Moreover, it has a small amount of calculation and better CPR artifact suppression performance.

[0120] Based on the above method, as Figure 2 shown, the present invention also provides a CPR artifact suppression system, including an acquisition unit, a preprocessing unit, a power spectral density calculation unit, a phase reconstruction unit, an adaptive filter construction unit, and an output unit.

[0121] The output end of the acquisition unit is connected to the input end of the preprocessing unit, the output end of the preprocessing unit is connected to the input end of the power spectral density calculation unit, the output end of the power spectral density calculation unit is connected to the input end of the phase reconstruction unit, the output end of the phase reconstruction unit is connected to the input end of the adaptive filter construction unit, the output end of the adaptive filter construction unit is connected to the input end of the output unit, and the output end of the output unit outputs the ECG data without CPR artifacts.

[0122] The acquisition unit is used to obtain the original ECG (electrocardiogram) data containing CPR (cardiopulmonary resuscitation) artifacts to be processed from existing databases (MIT and CU international authoritative electrocardiogram databases);

[0123] The preprocessing unit is used to preprocess the original ECG data and perform the first separation of the ECG signal and the CPR artifacts to obtain the CPR artifact signal without the ECG signal;

[0124] The power spectral density calculation unit is used to calculate the power spectral density corresponding to the CPR artifact signal;

[0125] The phase reconstruction unit is used to calculate the component frequency points of the CPR artifact signal according to the power spectral density, and then reconstruct the CPR artifact signal according to the component frequency points to obtain the CPR interference signal;

[0126] The adaptive filter construction unit is used to construct a Kalman adaptive filter to recursively solve the CPR interference signal to obtain the CPR artifact estimation;

[0127] An output unit, configured to remove CPR artifacts from the original ECG data according to the CPR artifact estimation amount and output ECG data without CPR artifacts.

[0128] The present invention further provides an electronic device, which includes a processor configured to run a computer program stored in a memory, so that the electronic device implements the steps of the method in the above embodiments.

[0129] The present invention further provides a computer-readable storage medium, in which a computer program is stored, and the computer program implements the steps of the method in the above embodiments when running on a processor.

[0130] The computer program includes computer program code, which may be in the form of source code, object code, executable file or some intermediate form, etc. The computer-readable medium may at least include: any entity or device capable of carrying the computer program code to an electronic device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disc, etc. In some jurisdictions, according to legislation and patent practice, the computer-readable medium may not be an electrical carrier signal and a telecommunication signal.

[0131] Those of ordinary skill in the art can understand that the above embodiments are specific embodiments for implementing the present invention, and in practical applications, various changes may be made in form and details without departing from the spirit and scope of the present invention.

Claims

1. A method for suppressing CPR artifacts, characterized in that, It includes the following steps: S1: Obtain the original ECG data containing CPR artifacts to be processed; S2: Preprocess the original ECG data, separate the ECG signal and the CPR artifacts for the first time, and obtain the CPR artifact signal without the ECG signal; S3: Calculate the power spectral density of the CPR artifact signal in S2; S4: Find the component frequency points of the CPR artifact signal according to the power spectral density PSD, and then reconstruct the CPR artifact signal according to the component frequency points to obtain the CPR interference signal; The S4 includes: S41: Calculate the maximum component frequency point of the CPR artifact signal according to the power spectral density: f0 = getMaxPos(PSD) (4) In formula (4), f0 represents the maximum component frequency point of the CPR artifact signal, that is, the first main frequency point; getMax represents the function to return the maximum value; getMaxPos represents the subscript of the maximum value of the power spectral density and the frequency point when the power spectral density is the maximum; S42: Reconstruct the CPR artifact signal according to the component frequency points to obtain the CPR interference signal: S cpr [n] = {A0[n] cos(ω0n + θ0) + K * A1[n] * cos(2ω0n + θ1)}; ω0 = 2πf0 (5) In formula (5), S cpr [n] represents the CPR interference signal after reconstruction; A0[n] represents the amplitude coefficient of the sine function constructed by the first main frequency point; ω0 represents the angular velocity; n represents the total amount of signals; θ0 represents the initial phase of the sine function constructed by the first main frequency point; A1[n] represents the amplitude coefficient of the sine function constructed by the second main frequency point; θ1 represents the initial phase of the sine function constructed by the second main frequency point; f0 represents the maximum component frequency point of the CPR artifact signal; K represents the coefficient for controlling the second main frequency point: K = 1; if f0 < f c K = 0; if f0 ≥ f c (6) In formula (6), f c represents the cut-off frequency; S5: Use a Kalman adaptive filter to recursively solve the CPR interference signal to obtain the CPR artifact estimator; S6: Remove the CPR artifact estimator from the original ECG data to obtain the ECG data without CPR artifacts.

2. The CPR artifact suppression method according to claim 1, wherein In the S1, the type of the original ECG data includes at least two waveforms; each waveform includes at least 100 samples, and each sample has at least 1000 sampling points.

3. The CPR artifact suppression method according to claim 1, wherein In the S2, the preprocessing method is: Use a Butterworth first-order filter to design a band-pass filter with a band-pass frequency of [1Hz, 3Hz], perform low-pass filtering on the original ECG data to remove the ECG signal therein, and realize the first separation of the ECG signal and the CPR artifacts.

4. The CPR artifact suppression method according to claim 1, wherein The S3 includes: Step S31: Perform autocorrelation calculation on the CPR artifact signal; In formula (1), R x (τ) represents the autocorrelation result of autocorrelation calculation on x(t); t represents time; τ represents time delay; x(t) represents the time-domain signal at time t after passing through the band-pass filter; x(t + τ) represents the time-domain signal at time t + τ after passing through the band-pass filter; Step S32: Perform a Fourier transform on the autocorrelation result R x (τ): In formula (2), S x (f) represents the Fourier transform result; j represents the imaginary part of a complex number; f represents the signal sampling frequency; τ represents the time delay; Step S33: Integrate the Fourier transform result S x (f) to obtain the power spectral density: In formula (3), PSD represents the power spectral density; S x (f) represents the result of Fourier transform.

5. The CPR artifact suppression method according to claim 1, wherein The S5 includes: S51: Establish a Kalman adaptive filter, and its model function is: x[n + 1] = φ[n + 1, n] * x[n] + v1[n] y[n] = C[n] * x[n] + v2[n] (7) In formula (7), x[n + 1] represents the state parameter at the (n + 1)-th moment; φ[n + 1, n] represents the Kalman state transition matrix; x[n] represents the state parameter at the n-th moment; v1[n] represents white noise with a mean of 0; y[n] represents the observed quantity, that is, the original ECG signal; C[n] represents the state transition matrix from x[n] to y[n]; v2[n] represents white noise with a mean of 0; x[n] = (A0[n] cosθ0[n], A0[n] sinθ0[n], A1[n] cosθ1[n], A1[n] sinθ1[n]) T ; C[n] = (cos(ω0n), -sin(ω0n), Kcos(2ω0n), -Ksin(2ω0n)) (8) In formula (8), x[n] represents the state parameter at time n; A0[n] represents the amplitude coefficient of the sine function constructed by the first main frequency point; θ0[n] represents the initial phase of the sine function constructed by the first main frequency point; A1[n] represents the amplitude coefficient of the sine function constructed by the second main frequency point; θ1[n] represents the initial phase of the sine function constructed by the second main frequency point; T represents the transpose matrix; ω0 represents the angular velocity; Step S52: By recursively solving the model function of the Kalman adaptive filter, the CPR artifact estimator S cpr [n]: S cpr [n] = C[n] * x[n].

6. The CPR artifact suppression method according to claim 1, wherein In S6, the method for obtaining ECG data without CPR artifacts is as follows: original ECG = y[n] - S cpr [n] (9) In Equation (9), origin ECG represents the ECG data without CPR artifacts; y[n] represents the observable quantity, that is, the original ECG data; S cpr [n] represents the CPR artifact estimator.

7. A CPR artifact suppression system based on the method according to any one of claims 1-6, characterized in that, It includes: An acquisition unit for obtaining the original ECG data containing CPR artifacts to be processed; A preprocessing unit for preprocessing the original ECG data to obtain a CPR artifact signal that does not contain the ECG signal; A power spectral density calculation unit for calculating the power spectral density corresponding to the CPR artifact signal; A phase reconstruction unit for calculating the component frequency points of the CPR artifact signal according to the power spectral density, and then reconstructing the CPR artifact signal according to the component frequency points to obtain a CPR interference signal; It includes the following steps: Calculating the maximum component frequency point of the CPR artifact signal according to the power spectral density: f0 = getMaxPos(PSD) (10) In formula (10), f0 represents the maximum component frequency point of the CPR artifact signal, that is, the first main frequency point; getMax represents the function for returning the maximum value; getMaxPos represents the subscript for returning the maximum value of the power spectral density and the frequency point when the power spectral density is the maximum; Reconstructing the CPR artifact signal according to the component frequency points to obtain a CPR interference signal: S cpr [n] = {A0[n] cos(ω0n + θ0) + K * A1[n] * cos(2ω0n + θ1)}; ω0 = 2πf0 (11) In Formula (11), S cpr [n] represents the CPR interference signal after reconstruction; A0[n] represents the amplitude coefficient of the sine function constructed by the first main frequency point; ω0 represents the angular velocity; n represents the total amount of signals; θ0 represents the initial phase of the sine function constructed by the first main frequency point; A1[n] represents the amplitude coefficient of the sine function constructed by the second main frequency point; θ1 represents the initial phase of the sine function constructed by the second main frequency point; f0 represents the maximum component frequency point of the CPR artifact signal; K represents the coefficient for controlling the second main frequency point: K = 1; if f0 < f c K = 0; if f0 ≥ f c (12) In formula (12), f c represents the cut-off frequency; An adaptive filter construction unit for constructing a Kalman adaptive filter to recursively solve the CPR interference signal to obtain a CPR artifact estimator; An output unit for removing the CPR artifact from the original ECG data according to the CPR artifact estimator and outputting the ECG data without CPR artifacts.

8. An electronic device, characterized in that, The electronic device includes a processor, and the processor is used to run the computer program stored in the memory so that the electronic device implements the steps of the method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and when the computer program runs on the processor, it implements the steps of the method according to any one of claims 1-6.

Images