CN110367936B

CN110367936B - Electrocardiosignal detection method and device

Info

Publication number: CN110367936B
Application number: CN201910718342.XA
Authority: CN
Inventors: 胡静
Original assignee: Guangzhou Shiyuan Electronics Thecnology Co Ltd
Current assignee: Guangzhou Shiyuan Electronics Thecnology Co Ltd
Priority date: 2019-08-05
Filing date: 2019-08-05
Publication date: 2022-01-28
Anticipated expiration: 2039-08-05
Also published as: CN110367936A

Abstract

The invention discloses an electrocardiosignal detection method and device. Wherein, the method comprises the following steps: acquiring electrocardiogram data corresponding to a plurality of leads to obtain waveform characteristics corresponding to the electrocardiogram data; decomposing the electrocardio data, and determining at least one target wavelet coefficient according to the processing result; respectively detecting at least one target wavelet coefficient and waveform characteristics based on a plurality of self-adaptive threshold detectors to obtain a plurality of detection results; and performing heart beat type detection on the electrocardiogram data and a plurality of detection results based on a preset network model to obtain a target detection result, wherein the target detection result represents whether the heart beat type corresponding to the electrocardiogram data is a bundle branch block. The invention solves the technical problem of low accuracy of the bundle branch block detection in the related technology.

Description

Electrocardiosignal detection method and device

Technical Field

The invention relates to the field of signal detection, in particular to an electrocardiosignal detection method and device.

Background

Right Bundle Branch Block (RBBB) is a type of cardiac electrical conduction system blocking disease. The right bundle branch of the heart is blocked from conduction, so that electrical signals cannot pass into the right ventricle via the right bundle branch, and must be activated by signals from the left ventricle, resulting in the occurrence of the disease. In which fig. 1 shows the process of cardiac electrical conduction, in fig. 1, the HIS Bundle (Atrio venticular Bundle) is called HIS Bundle, a special cardiac muscle of a mammalian heart, and is a part of the stimulation conduction system, which can transmit the excitation of the atrioventricular Node (i.e., the AV Node in fig. 1) to Purkinje fibers (e.g., Purkinje fibers in fig. 1). In fig. 1, laf (left analog fastcle) is the left front branch, lpf (left spatial fastcle) is the left interval branch, lbb (left Bundle branch) is the left Bundle branch, and rbb (right Bundle branch) is the right Bundle branch.

Currently, diseases that cause RBBB include at least: atrial septal defect, brucella syndrome, right ventricular hypertrophy, pulmonary embolism, ischemic heart disease, rheumatic fever, myocarditis, cardiomyopathy, hypertension, or the like. The basic pathophysiological defect in right bundle branch block is mainly due to the non-conduction of the electrical impulse from the HIS bundle to the right bundle branch, where the left bundle branch is normally conducted, and the right ventricular depolarization is significantly inconsistent with the left ventricle. This ventricular depolarization mismatch causes a change in the Electrocardiogram (ECG), and thus the right bundle branch block can be identified by extracting and analyzing the ECG features.

In the prior art, in the process of analyzing the electrocardiogram characteristics, all the extracted electrocardiogram characteristics are generally input into the same detector, and the detector outputs a detection result representing whether the patient has the RBBB disease. In addition, when acquiring the electrocardiographic signals, a plurality of leads are usually used to acquire the electrocardiographic signals, and the connection of the plurality of leads undoubtedly increases the complexity of the beam branch block detection.

In view of the above problems, no effective solution has been proposed.

Disclosure of Invention

The embodiment of the invention provides an electrocardiosignal detection method and device, which at least solve the technical problem of low accuracy of beam branch block detection in the related technology.

According to an aspect of the embodiments of the present invention, there is provided an electrocardiographic signal detection method, including: acquiring electrocardiogram data corresponding to a plurality of leads to obtain waveform characteristics corresponding to the electrocardiogram data; decomposing the electrocardio data, and determining at least one target wavelet coefficient according to the processing result; respectively detecting at least one target wavelet coefficient and waveform characteristics based on a plurality of self-adaptive threshold detectors to obtain a plurality of detection results; and performing heart beat type detection on the electrocardiogram data and a plurality of detection results based on a preset network model to obtain a target detection result, wherein the target detection result represents whether the heart beat type corresponding to the electrocardiogram data is a bundle branch block.

Further, the plurality of adaptive threshold detectors includes at least: the electrocardiosignal detection method comprises a first detector based on wavelet coefficients, a second detector based on waveform characteristics, a third detector based on waveform characteristics and a fourth detector based on waveform characteristics, wherein the electrocardiosignal detection method further comprises the following steps: detecting at least one target wavelet coefficient based on a first detector to obtain a first detection result; detecting the waveform characteristics based on a second detector to obtain a second detection result; detecting the waveform characteristics based on a third detector to obtain a third detection result; and detecting the waveform characteristics based on the fourth detector to obtain a fourth detection result.

Further, the electrocardiosignal detection method further comprises the following steps: acquiring wavelet coefficient sums and wavelet coefficient products corresponding to each lead; obtaining a first threshold corresponding to each lead according to the wavelet coefficient sum and the wavelet coefficient product; averaging first threshold values corresponding to the plurality of leads to obtain a first average value; obtaining a second threshold corresponding to each lead at the current moment according to the first threshold and the first parameter; and obtaining a first detection result according to the second threshold and the first average value, wherein the first detection result represents whether the cardiac beat type of the electrocardiographic data of each lead is bundle branch block or not.

Further, the electrocardiosignal detection method further comprises the following steps: decomposing the electrocardiogram data to obtain a preset number of wavelet coefficients, wherein the preset number of wavelet coefficients at least comprises: the second wavelet coefficient, the third wavelet coefficient, the fourth wavelet coefficient, the fifth wavelet coefficient, the sixth wavelet coefficient and the seventh wavelet coefficient, and the target wavelet coefficient at least comprises: a third wavelet coefficient, a fourth wavelet coefficient, a fifth wavelet coefficient; summing the third wavelet coefficient, the fourth wavelet coefficient, the fifth wavelet coefficient and the sixth wavelet coefficient to obtain a wavelet coefficient sum; summing the second wavelet coefficient, the sixth wavelet coefficient and the seventh wavelet coefficient to obtain a summation result; and performing product operation on the third wavelet coefficient and the summation result to obtain a wavelet coefficient product.

Further, the waveform characteristics include at least: the QRS wave duration and electrocardiosignal detection method further comprises the following steps: obtaining an S wave sequence and a Q wave sequence of waveform characteristics corresponding to each lead; obtaining a first duration of a QRS wave corresponding to each lead according to the S wave sequence and the Q wave sequence; averaging the first duration corresponding to the plurality of leads to obtain a second average value; obtaining a second duration corresponding to each lead according to the first duration and a second parameter; and obtaining a second detection result according to the second duration and the second average value, wherein the second detection result represents whether the cardiac beat type of the electrocardiographic data of each lead is the bundle branch block.

Further, the waveform characteristics include at least: the R wave deviation value and the electrocardiosignal detection method further comprise the following steps: obtaining a first R wave deviation value corresponding to each lead according to the electrocardiogram data corresponding to each lead; averaging the first R wave skewness values corresponding to the multiple leads to obtain a third average value; obtaining a second R wave deviation value corresponding to each R sequence according to the first R wave deviation value and a third parameter; and obtaining a third detection result according to the second R wave deviation value and the third average value, wherein the third detection result represents whether the cardiac beat type of the electrocardio data of each lead is the bundle branch block.

Further, the electrocardiosignal detection method further comprises the following steps: obtaining an R wave sequence in the electrocardiogram data corresponding to each lead; calculating a kernel density estimation value vector corresponding to the R wave sequence; and calculating the sequence length and the deviation value of the R wave sequence according to the kernel density estimated value vector.

Further, the waveform characteristics include at least: the P wave variance value and the electrocardiosignal detection method further comprise the following steps: obtaining a first P wave variation value corresponding to each lead according to the electrocardiogram data corresponding to each lead; averaging the first P wave variation values corresponding to the multiple leads to obtain a fourth average value; obtaining a second P wave variation value corresponding to each lead at the current moment according to the first P wave variation value and a fourth parameter; and obtaining a fourth detection result according to the second P wave variance value and the fourth average value, wherein the fourth detection result represents whether the cardiac beat type of the electrocardiographic data of each lead is the bundle branch block.

Further, the electrocardiosignal detection method further comprises the following steps: preprocessing the electrocardiogram data to obtain preprocessed electrocardiogram data; performing discrete wavelet decomposition on the preprocessed electrocardiogram data to obtain a plurality of wavelet coefficients; at least one target wavelet coefficient is determined from the plurality of wavelet coefficients.

Further, the electrocardiosignal detection method further comprises the following steps: performing first processing on the electrocardiogram data to obtain first electrocardiogram data, wherein the first processing at least comprises one of the following steps: impedance matching processing, filtering processing and amplifying processing; and carrying out second processing on the first electrocardiogram data to obtain second electrocardiogram data, wherein the second processing at least comprises the following steps: analog-to-digital conversion processing; and carrying out low-pass filtering processing on the second electrocardiogram data to obtain preprocessed electrocardiogram data.

According to another aspect of the embodiments of the present invention, there is also provided an electrocardiographic signal detection method, including: acquiring electrocardiogram data corresponding to a plurality of leads; the electrocardio data are respectively detected based on a plurality of self-adaptive threshold detectors to obtain a plurality of detection results; and performing heart beat type detection on the electrocardiogram data and a plurality of detection results based on a preset network model to obtain a target detection result, wherein the target detection result represents whether the heart beat type corresponding to the electrocardiogram data is a bundle branch block.

According to another aspect of the embodiments of the present invention, there is also provided an electrocardiographic signal detection apparatus, including: the acquisition module is used for acquiring the electrocardiogram data corresponding to the leads to obtain waveform characteristics corresponding to the electrocardiogram data; the decomposition module is used for decomposing the electrocardio data and determining at least one target wavelet coefficient according to the processing result; the first detection module is used for respectively detecting at least one target wavelet coefficient and waveform characteristics based on a plurality of self-adaptive threshold detectors to obtain a plurality of detection results; and the second detection module is used for carrying out heart beat type detection on the electrocardio data and the plurality of detection results based on the preset network model to obtain a target detection result, wherein the target detection result represents whether the heart beat type corresponding to the electrocardio data is a bundle branch block.

According to another aspect of the embodiments of the present invention, there is also provided a storage medium including a stored program, wherein when the program runs, the apparatus on which the storage medium is located is controlled to execute the above-mentioned electrocardiographic signal detection method.

According to another aspect of the embodiments of the present invention, there is also provided a processor, configured to execute a program, where the program executes the above-mentioned cardiac signal detection method.

In the embodiment of the invention, a mode that a plurality of detectors respectively detect waveform characteristics is adopted, after electrocardiographic data corresponding to a plurality of leads is obtained, corresponding waveform characteristics are extracted from the electrocardiographic data, the electrocardiographic data is decomposed, at least one target wavelet coefficient is determined according to a processing result, then the at least one target wavelet coefficient and the waveform characteristics are respectively detected based on a plurality of adaptive threshold detectors to obtain a plurality of detection results, and finally, the electrocardiographic data and the plurality of detection results are subjected to heartbeat type detection based on a preset network model to obtain a target detection result.

In the process, different adaptive threshold detectors are adopted to respectively carry out feature analysis on the target wavelet coefficient and the waveform feature, so that the analysis of the target wavelet coefficient and the waveform feature is more targeted, and the detection result output by the adaptive threshold detectors is more accurate, so that the target detection result is more accurate and the stability is higher.

Therefore, the scheme provided by the application achieves the purpose of detecting whether the cardiac beat type corresponding to the electrocardiogram data is the bundle branch block, thereby realizing the technical effect of improving the accuracy of bundle branch block detection and further solving the technical problem of low accuracy of bundle branch block detection in the related technology.

Drawings

The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the invention without limiting the invention. In the drawings:

FIG. 1 is a schematic illustration of a process of cardiac electrical conduction according to the prior art;

FIG. 2 is a flow chart of a method for detecting an ECG signal according to an embodiment of the invention;

FIG. 3 is a waveform diagram corresponding to an alternative right bundle branch block in accordance with embodiments of the present invention;

FIG. 4 is a schematic illustration of alternative electrocardiographic data in accordance with embodiments of the present invention;

fig. 5 is a schematic diagram of an alternative QRS complex in accordance with an embodiment of the present invention;

FIG. 6 is a schematic diagram of an alternative deep learning network in accordance with embodiments of the present invention;

FIG. 7 is a flow chart of a method of detecting an ECG signal according to an embodiment of the invention; and

fig. 8 is a schematic diagram of an electrocardiograph signal detection apparatus according to an embodiment of the present invention.

Detailed Description

In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

It should be noted that the terms "first," "second," and the like in the description and claims of the present invention and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

Example 1

In accordance with an embodiment of the present invention, there is provided an embodiment of a method for cardiac electrical signal detection, it should be noted that the steps illustrated in the flowchart of the accompanying drawings may be executed in a computer system, such as a set of computer executable instructions, and that while a logical order is illustrated in the flowchart, in some cases, the steps illustrated or described may be executed in an order different than that illustrated or described herein.

Fig. 2 is a flowchart of a method for detecting an ecg signal according to an embodiment of the present invention, as shown in fig. 2, the method includes the following steps:

step S202, obtaining the electrocardiogram data corresponding to a plurality of leads to obtain the waveform characteristics corresponding to the electrocardiogram data.

In step S202, the lead is one of the terms of electrocardiogram, and refers to the placement of the electrodes on the body surface and the connection between the electrodes and the amplifier when recording the electrocardiogram. Optionally, each lead acquires corresponding electrocardiographic data, and preferably, the leads at least include a V1-V3 lead, that is, the electrocardiographic data corresponding to the leads in this application is electrocardiographic data corresponding to a V1-V3 lead.

It should be noted that for right bundle branch block, the electrical signal from the left ventricle must be transmitted through the myocardium, and the transmission speed of this route is slower than that of the original his bundle-purkinje fiber route, so on the electrocardiogram, the QRS complex of the patient is wider, wherein the QRS wave is the largest amplitude complex in the normal electrocardiogram, which reflects the whole process of ventricular depolarization. For right bundle branch block, the patient ' S ventricular rhythm must come from above the ventricle (i.e., sinoatrial node, atrial, or atrioventricular node), the duration of propagation of the QRS complex wave is longer than 100 milliseconds in the case of incomplete block, or longer than 120 milliseconds in the case of complete block, the V1 lead has an end-stage R wave (e.g., R, rR ', rsR ', rSR ', or qR, etc. waveforms), there is an RSR ' pattern (e.g., ' M-shaped ' QRS complex) in the V1 to V3 leads, the waveform diagram on the electrocardiogram corresponding to the right bundle branch block shown in fig. 3 shows a double R wave for the V1 lead, and its corresponding S wave is wider and ambiguous for the V6 lead. The above is the waveform characteristics of the electrocardiogram corresponding to the right bundle branch block.

From the waveform characteristics of the right bundle branch block, it can be seen that the supraventricular rhythm and the QRS complex time length can be observed on all leads, while the more specific 'M-shaped' QRS complex can be observed only on the V1-V3 leads. Therefore, only the V1-V3 lead need be analyzed.

And step S204, decomposing the electrocardio data, and determining at least one target wavelet coefficient according to the processing result.

In step S204, the decomposition processing performed on the electrocardiographic data may be discrete wavelet decomposition. Specifically, the method comprises the steps of preprocessing electrocardiographic data to obtain preprocessed electrocardiographic data, then performing discrete wavelet decomposition on the preprocessed electrocardiographic data to obtain a plurality of wavelet coefficients, and finally determining at least one target wavelet coefficient from the plurality of wavelet coefficients.

It should be noted that, in the present application, multichannel synchronous data is used to acquire electrocardiographic data to be processed, and at this time, the acquired electrocardiographic data includes cardiac signals, background noise, and electrocardiographic data to be detected. Therefore, before detecting the electrocardiographic data, the electrocardiographic data needs to be preprocessed to filter out the miscellaneous signals (i.e. the human heart signals, the background noise, etc.) in the electrocardiographic data.

Optionally, first processing is performed on the electrocardiographic data to obtain first electrocardiographic data, then second processing is performed on the first electrocardiographic data to obtain second electrocardiographic data, and finally low-pass filtering processing is performed on the second electrocardiographic data to obtain preprocessed electrocardiographic data. Wherein the first processing includes at least one of: impedance matching processing, filtering processing and amplifying processing, wherein the second processing at least comprises the following steps: and D, analog-to-digital conversion processing.

Specifically, after acquiring electrocardiographic data including a human heart signal and background noise by using multichannel synchronous data, first, electrocardiographic data (that is, electrocardiographic data including a human heart signal and background noise) is acquired through a lead and a sensor, and the acquired electrocardiographic data is subjected to processing (that is, first processing) such as impedance matching, filtering, amplification and the like by using an analog circuit. Then, the analog signal corresponding to the electrocardiographic data is converted into a digital signal by the analog-to-digital converter (i.e., the second processing), and the digital signal is stored in the memory. The actually acquired electrocardiogram data contains various noises, and the waveform is rough and unsmooth, so that the useful information contained in the QRS complex wave is difficult to extract. At this time, low-pass filtering is further performed using a low-pass digital filter (e.g., a butterworth filter) to filter out high-frequency noise (e.g., noise having a frequency of 300Hz or higher) to obtain filtered electrocardiographic data o (t). Optionally, fig. 4 shows a schematic of the preprocessed electrocardiographic data containing bundle branch blocks. In fig. 4, I, II and III represent three different waveform sequences, respectively, and VR, VL, VP, V1, V2, V3, V4, V5, and V6 represent different leads, respectively.

Further, in the present application, Daubechies4(Db4) wavelet is used to perform discrete wavelet decomposition on the preprocessed electrocardiographic data, where the number of decomposition layers is N, and in this embodiment, when N is 7, the practical performance is the best, and the wavelet decomposition coefficients D1, D2, D3, D4, D5, D6, D7, and a7 are obtained, so that the number of wavelet decomposition layers in the present application is preferably 7. In addition, since the wavelet coefficients D3, D4 and D5 contain the maximum information and the maximum energy, the QRS complex is detected by using the wavelet coefficients D3, D4 and D5 as the target wavelet coefficients, and the detected QRS complex is as shown in fig. 5, three connected waves of Q wave, R wave and S wave form the QRS complex, and the P wave, T wave and U wave do not belong to the QRS complex.

Step S206, based on the adaptive threshold detectors, at least one target wavelet coefficient and waveform characteristics are detected respectively, and a plurality of detection results are obtained.

In step S206, the plurality of adaptive threshold detectors includes at least: the wavelet feature-based detection device comprises a first detector based on wavelet coefficients, a second detector based on waveform features, a third detector based on waveform features and a fourth detector based on waveform features, wherein at least one target wavelet coefficient is detected based on the first detector to obtain a first detection result; detecting the waveform characteristics based on a second detector to obtain a second detection result; detecting the waveform characteristics based on a third detector to obtain a third detection result; and detecting the waveform characteristics based on the fourth detector to obtain a fourth detection result.

It should be noted that in step S206, different waveform characteristics correspond to different adaptive threshold detectors, wherein the second detector is a detector based on the duration of QRS wave, the third detector is a detector based on the probability density distribution of R wave, and the fourth detector is a detector based on the variability of P wave.

Through step S206, different detectors are used to detect different waveform characteristics and wavelet coefficients, so that analysis of the target wavelet coefficients and the waveform characteristics is more targeted, and the detection result of the detectors is improved.

And S208, performing heart beat type detection on the electrocardio data and the plurality of detection results based on a preset network model to obtain a target detection result, wherein the target detection result represents whether the heart beat type corresponding to the electrocardio data is a bundle branch block.

In step S208, the heartbeat corresponding to the electrocardiographic data is a single heartbeat, which is more suitable for practical application, so that the application range of the present application can be expanded by using the single heartbeat. The preset network model may be a deep learning network.

Optionally, after a detection result obtained by each adaptive threshold detector is obtained, the detection result and filtered and extracted electrocardiographic data are input into a deep learning Network RBNN (Radial Basis Neural Network) together, and after convolution operation, a target detection result is obtained. First, a deep learning network is constructed, as shown in FIG. 6, to detect right bundle branch block by using RBNN, which is composed of a residual convolutional network (ResNet) and a Bi-directional long-term short-term cyclic neural network (Bi-LSTM). Fig. 6 shows the structure of the RBNN described above, in which the RBNN inputs electrocardiographic data and detection results detected by a plurality of adaptive threshold detectors, and the output is a target detection result characterizing the heartbeat type (RBBB or N-RBBB). The residual convolutional network is located on top of the RBNN network model. After the residual convolutional network, the Bi-LSTM layer, the Flatten layer, and the fully-connected layer are placed in order. For the residual convolutional network, the convolutional kernel size and number of channels are set to 11 and 64, respectively, and the number of neurons in the Bi-LSTM layer and the fully-connected layer are 32 and 64, respectively. And obtaining a detection result of the right bundle branch block through the RBNN model.

It should be noted that the detection result of the bundle branch block can be displayed on a single lead electrocardiogram plaster, a multi-modality device and a monitor device which comprise an electrocardiogram module, and is used as the basis for detection and diagnosis of individuals or doctors. In addition, in fig. 6, the convolution function is 1 dconvolume (128, 11); the Batch Normalization function is Batch Normalization, which is used to unify scattered data; a Linear rectification function (Rectified Linear Unit, called as a modified Linear Unit), is a commonly used activation function (activation function) in an artificial neural network, and generally refers to a nonlinear function represented by a ramp function and a variation thereof; the maximum pooling function is MaxPool, and the sampling kernel is 2; the sampling function is a downsample function which can extract original data through downsampling; the Flatten layer is used for flattening input, namely, the multidimensional input is subjected to one-dimensional input, and is usually used for transition from a convolution layer to a full-connection layer; dropout is the probability of discarding the neural network unit from the network temporarily according to a certain probability in the training process of the deep learning network, and the probability of discarding is 0.2 in fig. 6.

Based on the schemes defined in steps S202 to S208, it can be known that, after acquiring electrocardiographic data corresponding to a plurality of leads, a manner of detecting waveform characteristics by a plurality of detectors is adopted, corresponding waveform characteristics are extracted from the electrocardiographic data, decomposition processing is performed on the electrocardiographic data, at least one target wavelet coefficient is determined according to a processing result, then, the at least one target wavelet coefficient and the waveform characteristics are detected by the plurality of adaptive threshold detectors, a plurality of detection results are obtained, and finally, cardiac beat type detection is performed on the electrocardiographic data and the plurality of detection results based on a preset network model, so as to obtain a target detection result.

It is easy to notice that in the above process, different adaptive threshold detectors are adopted to perform feature analysis on the target wavelet coefficient and the waveform feature respectively, so that the analysis of the target wavelet coefficient and the waveform feature is more targeted, and the detection result output by the adaptive threshold detector is more accurate, so that the target detection result is more accurate and the stability is higher.

In an alternative embodiment, the detecting at least one target wavelet coefficient based on the first detector to obtain the first detection result includes: the method comprises the steps of obtaining wavelet coefficient sums and wavelet coefficient products corresponding to each lead, obtaining a first threshold corresponding to each lead according to the wavelet coefficient sums and the wavelet coefficient products, averaging the first thresholds corresponding to a plurality of leads to obtain a first average value, obtaining a second threshold corresponding to each lead at the current moment according to the first thresholds and first parameters, and finally obtaining a first detection result according to the second thresholds and the first average value, wherein the first detection result represents whether the heart beat type of the electrocardiograph data of each lead is bundle branch block or not.

It should be noted that before the detection of the wavelet coefficients based on the first detector, the wavelet coefficient threshold is determined first, and in this embodiment, the wavelet coefficient threshold is determined based on the V1-V3 lead, that is, the first threshold and the second threshold are determined. Wherein the first threshold and the second threshold are related to a wavelet coefficient sum and a wavelet coefficient product.

Optionally, the electrocardiographic data is decomposed to obtain a preset number of wavelet coefficients, where the preset number of wavelet coefficients at least includes: the second wavelet coefficient, the third wavelet coefficient, the fourth wavelet coefficient, the fifth wavelet coefficient, the sixth wavelet coefficient and the seventh wavelet coefficient, and the target wavelet coefficient at least comprises: a third wavelet coefficient, a fourth wavelet coefficient, a fifth wavelet coefficient; and finally, performing product operation on the third wavelet coefficient and the summation result to obtain a wavelet coefficient product.

Wherein the wavelet coefficients and X satisfy the following formula:

X＝D3+D4+D5+D6

in the above equation, D3 is the third wavelet coefficient, D4 is the fourth wavelet coefficient, D5 is the fifth wavelet coefficient, and D6 is the sixth wavelet coefficient.

The wavelet coefficient product Y satisfies the following formula:

in the above equation, D2 is the second wavelet coefficient, D7 is the seventh wavelet coefficient, D5 is the fifth wavelet coefficient, and D6 is the sixth wavelet coefficient.

After the wavelet coefficient sum and the wavelet coefficient product are obtained, a first threshold value Z corresponding to each lead is obtained according to the wavelet coefficient sum and the wavelet coefficient product, namely Z satisfies the following formula:

Z＝X×Y

optionally, for the electrocardiographic data of the V1-V3 lead, if the number of leads is M, then M is 3, at this time, M Z values may be extracted, and an average value Zmean (i.e., a first average value) thereof is calculated, where Zmean satisfies the following equation:

in the above formula, Zs represents the Z value of the s-th lead, and Zmean represents the Zmean value calculated at the q-th time in the continuous monitoring process. Wherein, the initial threshold value of Zmean is Zmean D0, Zmean Dq represents the qth Z threshold value, and the iteration update is carried out by the following formula to obtain the second threshold value:

ZmeanD_q+1＝λ₁ZmeanD_q+μ₁ZmeanD_q-1

in the above equation, ZmeanDq is the second threshold, and λ 1 and μ 1 are the first parameters, where λ 1+ μ 1 is 1. Preferably, in the present application, λ 1 is 0.85 and μ 1 is 0.15.

When the Zmeannq value is larger than ZmeanDq, the heart beat type of the electrocardio data of each lead is right bundle branch block, and conversely, the heart beat type of the electrocardio data of each lead is non-right bundle branch block.

In an alternative embodiment, the waveform characteristics include at least: the duration of the QRS wave, wherein the waveform feature is detected based on the second detector, and a second detection result is obtained, including: obtaining an S wave sequence and a Q wave sequence of waveform characteristics corresponding to each lead, obtaining a first duration of QRS waves corresponding to each lead according to the S wave sequence and the Q wave sequence, averaging the first durations corresponding to a plurality of leads to obtain a second average value, obtaining a second duration corresponding to each lead according to the first duration and a second parameter, and finally obtaining a second detection result according to the second duration and the second average value, wherein the second detection result represents whether the heart beat type of the electrocardiographic data of each lead is bundle branch block or not.

It should be noted that the waveform characteristics of the electrocardiographic data of the right bundle branch block on the V1-V3 lead have the characteristic that the duration of the QRS complex is more than 120 ms. Therefore, a second detection result can be obtained by extracting features related to the duration of the QRS wave (used for analyzing the duration of the QRS wave) from the electrocardio data and analyzing the features.

Optionally, X _ P, X _ Q, X _ R, X _ S and X _ T denote the sequence of P, Q, R, S and T waves, respectively, wherein the duration of the QRS complex can be expressed as QRS _ width, and the duration of the QRS complex satisfies the following equation:

QRS_width＝(X_S-X_Q)/fs

in the above equation, fs represents a sampling frequency.

Optionally, for the electrocardiographic data of the V1-V3 lead, if the number of leads is M, then M is 3, at this time, M QRS _ width values may be extracted, and an average value QRS _ width mean (i.e., a second average value) thereof is calculated, that is, QRS _ width mean satisfies the following formula:

in the above formula, QRS _ width represents the QRS _ width value of the s-th lead, QRS _ width means q represents the QRS _ width mean value calculated from the q-th sequence during continuous monitoring, wherein the initial threshold of QRS _ width mean is QRS _ width mean 0 (e.g., 120ms), QRS _ width mean dq represents the q-th QRS _ width mean threshold, and the second duration is iteratively updated by the following formula:

QRS_widthmeanD_q+1＝λ₂QRS_widthmeanD_q+μ₂QRS_widthmeanD_q-1

in the above equation, QRS _ width means dq is the second duration, and λ 2 and μ 2 are the second parameters, where λ 2+ μ 2 is 1. Preferably, in the present application, λ 2 is 0.85 and μ 2 is 0.15.

It should be noted that, based on the waveform characteristics corresponding to the right bundle branch block, the longer the QRS duration (QRS duration >120ms), the more likely it is the right bundle branch block. When the QRS _ width means q value obtained through real-time detection is larger than the QRS _ width means Dq value, the heart beat type of the electrocardio data of each lead is right bundle branch block, and conversely, the heart beat type of the electrocardio data of each lead is non-right bundle branch block.

In an alternative embodiment, the waveform characteristics include at least: r wave deviation value, wherein, based on the third detector detects the wave form characteristic, obtains the third testing result, includes: obtaining a first R wave deviation value corresponding to each lead according to the electrocardiogram data corresponding to each lead; averaging the first R wave skewness values corresponding to the multiple leads to obtain a third average value; obtaining a second R wave deviation value corresponding to each R sequence according to the first R wave deviation value and a third parameter; and obtaining a third detection result according to the second R wave deviation value and the third average value, wherein the third detection result represents whether the cardiac beat type of the electrocardio data of each lead is the bundle branch block.

The waveform of the electrocardiographic data of the right bundle branch block on the V1-V3 lead has the characteristics of an end R wave (for example, R, rR ', rsR ', rSR ', or qR and the like). Therefore, a probability density distribution including the R-wave (for analyzing whether the R-wave is monophasic or biphasic) is extracted from the electrocardiographic data, and analyzed to obtain a third detection result.

Optionally, obtaining a first R-wave skewness value corresponding to each lead according to the electrocardiographic data corresponding to each lead includes: obtaining an R wave sequence in the electrocardiogram data corresponding to each lead; calculating a kernel density estimation value vector corresponding to the R wave sequence; and calculating the sequence length and the deviation value of the R wave sequence according to the kernel density estimated value vector.

Specifically, X _ P, X _ Q, X _ R, X _ S and X _ T respectively represent sequences of P, Q, R, S waves and T waves, and probability density distribution of R waves is calculated, wherein ecg represents a sequence corresponding to electrocardiographic data, ksDensity is an operator for calculating sequence kernel density estimation and is used for calculating kernel density estimation of a sample observation vector X, xi is a vector formed by 100 points selected at equal intervals in a value range of X, and f1 is a kernel density estimation value vector corresponding to xi. The kernel function used is a Gaussian kernel function; skewness and kurtosis are operators for calculating sequence length, skewness and kurtosis respectively, and are shown as follows:

[f1,xi]＝ksdensity(X_R)

kurt_R＝kurtosis(f1)

skew_R＝skewness(f1)

based on the waveform characteristics, the greater the skewness of the R wave (i.e., the greater the skewness of the bi-phase wave), the more likely the waveforms such as R, rR ', rsR ', rSR ' or qR are to appear, i.e., the greater the probability of the right bundle branch block appearing. Therefore, for the electrocardiographic data of the V1-V3 lead, if the number of leads is M, then M is 3, in this case, M skew _ R values are extracted, and the average value skew _ Rmean (i.e. the third average value) is calculated as shown in the following formula:

in the above formula, skew _ Rs represents a skew _ R value of the s-th lead, and skew _ Rmeanq represents a skew _ Rmean value calculated q-th in the continuous monitoring process, wherein an initial threshold of the skew _ Rmean is skew _ RmeanD0, and the skew _ RmeanDq represents a q-th skew _ R threshold, and the skew _ RmeanDq is iteratively updated according to the following formula to obtain a second R wave skewness value:

skew_RmeanD_q+1＝λ₃skew_RmeanD_q+μ₃skew_RmeanD_q-1

in the above formula, skew _ RmeanDq is a second R-wave deviation value, and λ 3 and μ 3 are third parameters, where λ 3+ μ 3 is 1. Preferably, in the present application, λ 3 is 0.8 and μ 3 is 0.2.

It should be noted that when the value of skew _ Rmeanq detected in real time is greater than skew _ RmeanDq, the cardiac beat type of the electrocardiographic data of each lead is a right bundle branch block, and conversely, the cardiac beat type of the electrocardiographic data of each lead is a non-right bundle branch block.

In an alternative embodiment, the waveform characteristics include at least: the P-wave variance value, wherein the waveform feature is detected based on a fourth detector to obtain a fourth detection result, including: obtaining a first P wave variation value corresponding to each lead according to the electrocardiogram data corresponding to each lead; averaging the first P wave variation values corresponding to the multiple leads to obtain a fourth average value; obtaining a second P wave variation value corresponding to each lead at the current moment according to the first P wave variation value and a fourth parameter; and obtaining a fourth detection result according to the second P wave variance value and the fourth average value, wherein the fourth detection result represents whether the cardiac beat type of the electrocardiographic data of each lead is the bundle branch block.

It should be noted that the waveform characteristics of the electrocardiographic data of the right bundle branch block on the V1-V3 lead has the characteristics of the supraventricular rhythm. Among them, the main manifestation of supraventricular rhythm is normal P-waveform, and non-supraventricular rhythm is usually changed. Therefore, a fourth detection result can be obtained by extracting a P-wave morphology feature including the variability of the P-wave from the electrocardiographic data and analyzing the P-wave morphology feature.

Optionally, X _ P, X _ Q, X _ R, X _ S and X _ T denote sequences of P, Q, R, S and T waves, respectively, let X (n), n be 1. The phase space (y (n)) and P-wave variability (PIV) of the P-wave are calculated as follows:

y(n)＝(x(n),x(n+1),...,x(n+(m-1)t),n＝1,2,....,m

wherein, | | | represents euclidean distance, h is a step function, m is an embedding dimension, t is delay time, C is a combination operation, and r is a parameter.

For the V1-V3 lead electrocardiographic data, the number of leads is M, and M is 3, in this case, M PIV values can be extracted, and the average value PIVmean (i.e., the fourth average value) thereof is calculated:

in the above formula, PIVs represents the PIV value of the s-th lead, pivmanq represents the calculated PIVmean value of the q-th lead in the continuous monitoring process, wherein the initial threshold of PIVmean is pivmad 0, pivmandq represents the q-th PIV threshold, and the iterative update is performed by the following formula to obtain the second P-wave variance value:

PIVmeanD_q+1＝λ₄PIVmeanD_q+μ₄PIVmeanD_q-1

in the above formula, pivmaedq is the second P-wave variation value, and λ 4 and μ 4 are the fourth parameters, where λ 4+ μ 4 is 1. Preferably, in the present application, λ 4 is 0.85 and μ 4 is 0.15.

Further, after the four detection results are obtained, the four detection results and the electrocardiogram data are input into a preset network model for processing, and then a target detection result can be obtained.

According to the content, the scheme provided by the application is classified and analyzed based on the single-heart beat, and the requirement of practical use is better met; in addition, the characteristic analysis is carried out based on the waveform characteristics of the right bundle branch block, the accuracy rate and the stability are higher, and the practical application requirements are met. In addition, the scheme provided by the application also combines the traditional characteristics and the depth characteristics, the generalization capability of the model is stronger, and the practicability is wider due to the adoption of integration of a plurality of models.

Example 2

According to an embodiment of the present invention, an embodiment of an electrocardiograph signal detection method is further provided, and fig. 7 is a flowchart of the electrocardiograph signal detection method according to the embodiment of the present invention, as shown in fig. 7, the method includes the following steps:

step S702, acquiring electrocardiogram data corresponding to a plurality of leads.

In step S702, corresponding electrocardiographic data is acquired for each lead, preferably, the leads at least include the V1-V3 leads, that is, the electrocardiographic data corresponding to the leads in this application is electrocardiographic data corresponding to the V1-V3 leads.

In addition, after the electrocardio data are obtained, the electrocardio data are preprocessed to filter out miscellaneous signals in the electrocardio data. Specifically, first processing is performed on the electrocardiographic data to obtain first electrocardiographic data, then second processing is performed on the first electrocardiographic data to obtain second electrocardiographic data, and finally low-pass filtering processing is performed on the second electrocardiographic data to obtain preprocessed electrocardiographic data. Wherein the first processing includes at least one of: impedance matching processing, filtering processing and amplifying processing, wherein the second processing at least comprises the following steps: and D, analog-to-digital conversion processing.

Step S704, detecting the electrocardiographic data based on the plurality of adaptive threshold detectors, respectively, to obtain a plurality of detection results.

In step S704, each adaptive threshold detector processes a different feature of the electrocardiographic data.

Optionally, after obtaining the electrocardiographic data, performing feature extraction on the electrocardiographic data to obtain a waveform feature corresponding to the electrocardiographic data, where the waveform feature at least includes: duration of QRS wave, deviation value of R wave and variance value of P wave. And then processing the waveform characteristics based on different adaptive threshold detectors respectively to obtain detection results. The detection corresponding to the three waveform features may be the second detector, the third detector, and the fourth detector in embodiment 1, and the specific detection process is described in embodiment 1 and is not described herein again.

Through step S704, different waveform features are detected by different detectors, so that analysis of the waveform features is more targeted, and the detection result of the detectors is improved.

Step S706, performing heart beat type detection on the electrocardiographic data and the plurality of detection results based on a preset network model to obtain a target detection result, wherein the target detection result represents whether the heart beat type corresponding to the electrocardiographic data is a bundle branch block.

Based on the schemes defined in steps S702 to S706, it can be known that, after acquiring electrocardiographic data corresponding to a plurality of leads, extracting corresponding waveform features from the electrocardiographic data by using a manner that a plurality of detectors detect waveform features respectively, then detecting the waveform features based on a plurality of adaptive threshold detectors respectively to obtain a plurality of detection results, and finally performing heartbeat type detection on the electrocardiographic data and the plurality of detection results based on a preset network model to obtain a target detection result.

It is easy to notice that in the above process, different adaptive threshold detectors are adopted to perform feature analysis on the waveform features respectively, so that the analysis of the waveform features is more targeted, and the detection result output by the adaptive threshold detector is more accurate, so that the target detection result is more accurate and the stability is higher.

Example 3

According to an embodiment of the present invention, there is further provided an embodiment of an electrocardiographic signal detection apparatus, and fig. 8 is a schematic diagram of the electrocardiographic signal detection apparatus according to the embodiment of the present invention, as shown in fig. 8, the apparatus includes: an acquisition module 801, a decomposition module 803, a first detection module 805, and a second detection module 807.

The acquisition module 801 is used for acquiring electrocardiographic data corresponding to a plurality of leads to obtain waveform characteristics corresponding to the electrocardiographic data; the decomposition module 803 is configured to perform decomposition processing on the electrocardiographic data, and determine at least one target wavelet coefficient according to a processing result; a first detection module 805, configured to detect at least one target wavelet coefficient and waveform characteristics based on multiple adaptive threshold detectors, respectively, to obtain multiple detection results; the second detecting module 807 is configured to perform heartbeat type detection on the electrocardiographic data and the plurality of detection results based on the preset network model to obtain a target detection result, where the target detection result represents whether a heartbeat type corresponding to the electrocardiographic data is a bundle branch block.

It should be noted here that the acquiring module 801, the decomposing module 803, the first detecting module 805, and the second detecting module 807 correspond to steps S202 to S208 of the above embodiment, and the four modules are the same as the corresponding steps in the implementation example and application scenario, but are not limited to the disclosure in the above embodiment.

In an alternative embodiment, the plurality of adaptive threshold detectors comprises at least: the wavelet coefficient-based first detector, the waveform feature-based second detector, the waveform feature-based third detector and the waveform feature-based fourth detector, wherein the first detection module comprises: the device comprises a third detection module, a fourth detection module, a fifth detection module and a sixth detection module. The third detection module is used for detecting at least one target wavelet coefficient based on the first detector to obtain a first detection result; the fourth detection module is used for detecting the waveform characteristics based on the second detector to obtain a second detection result; the fifth detection module is used for detecting the waveform characteristics based on the third detector to obtain a third detection result; and the sixth detection module is used for detecting the waveform characteristics based on the fourth detector to obtain a fourth detection result.

In an alternative embodiment, the third detection module comprises: the device comprises a first acquisition module, a first processing module, a second processing module, a third processing module and a seventh detection module. The first acquisition module is used for acquiring the wavelet coefficient sum and the wavelet coefficient product corresponding to each lead; the first processing module is used for obtaining a first threshold value corresponding to each lead according to the wavelet coefficient sum and the wavelet coefficient product; the second processing module is used for averaging first thresholds corresponding to the plurality of leads to obtain a first average value; the third processing module is used for obtaining a second threshold value corresponding to each lead at the current moment according to the first threshold value and the first parameter; and the seventh detection module is used for obtaining a first detection result according to the second threshold and the first average value, wherein the first detection result represents whether the heart beat type of the electrocardiographic data of each lead is a bundle branch block.

In an alternative embodiment, the first obtaining module includes: the device comprises a fourth processing module, a fifth processing module, a sixth processing module and a seventh processing module. The fourth processing module is configured to perform decomposition processing on the electrocardiographic data to obtain a preset number of wavelet coefficients, where the preset number of wavelet coefficients at least includes: the second wavelet coefficient, the third wavelet coefficient, the fourth wavelet coefficient, the fifth wavelet coefficient, the sixth wavelet coefficient and the seventh wavelet coefficient, and the target wavelet coefficient at least comprises: a third wavelet coefficient, a fourth wavelet coefficient, a fifth wavelet coefficient; the fifth processing module is used for summing the third wavelet coefficient, the fourth wavelet coefficient, the fifth wavelet coefficient and the sixth wavelet coefficient to obtain a wavelet coefficient sum; the sixth processing module is used for summing the second wavelet coefficient, the sixth wavelet coefficient and the seventh wavelet coefficient to obtain a summation result; and the seventh processing module is used for performing product operation on the third wavelet coefficient and the summation result to obtain a wavelet coefficient product.

In an alternative embodiment, the waveform characteristics include at least: QRS wave duration, wherein the fourth detection module comprises: the device comprises a second obtaining module, an eighth processing module, a ninth processing module, a tenth processing module and an eleventh processing module. The second acquisition module is used for acquiring an S wave sequence and a Q wave sequence of waveform characteristics corresponding to each lead; the eighth processing module is used for obtaining the first duration of the QRS wave corresponding to each lead according to the S wave sequence and the Q wave sequence; the ninth processing module is used for averaging the first duration corresponding to the plurality of leads to obtain a second average value; the tenth processing module is used for obtaining a second duration corresponding to each lead according to the first duration and the second parameter; and the eleventh processing module is configured to obtain a second detection result according to the second duration and the second average value, where the second detection result represents whether the cardiac beat type of the electrocardiographic data of each lead is a bundle branch block.

In an alternative embodiment, the waveform characteristics include at least: an R-wave deviation value, wherein the fifth detection module comprises: the device comprises a third acquisition module, an eleventh processing module, a twelfth processing module and a thirteenth processing module. The third acquisition module is used for obtaining a first R wave deviation value corresponding to each lead according to the electrocardiogram data corresponding to each lead; the eleventh processing module is used for averaging the first R wave skewness values corresponding to the multiple leads to obtain a third average value; the twelfth processing module is used for obtaining a second R wave deviation value corresponding to each R sequence according to the first R wave deviation value and the third parameter; and the thirteenth processing module is used for obtaining a third detection result according to the second R-wave deviation value and the third average value, wherein the third detection result represents whether the cardiac beat type of the electrocardiographic data of each lead is a bundle branch block.

In an alternative embodiment, the third obtaining module includes: a fourth obtaining module, a fourteenth processing module and a fifteenth processing module. The fourth acquisition module is used for acquiring an R wave sequence in the electrocardiogram data corresponding to each lead; the fourteenth processing module is used for calculating a kernel density estimation value vector corresponding to the R wave sequence; and the fifteenth processing module is used for calculating the sequence length and the deviation value of the R wave sequence according to the kernel density estimated value vector.

In an alternative embodiment, the waveform characteristics include at least: p ripples variation value, wherein, the sixth detection module includes: the device comprises a fifth obtaining module, a sixteenth processing module, a seventeenth processing module and an eighteenth processing module. The fifth acquisition module is used for acquiring a first P wave variation value corresponding to each lead according to the electrocardiogram data corresponding to each lead; a sixteenth processing module, configured to average first P-wave variation values corresponding to the multiple leads to obtain a fourth average value; a seventeenth processing module, configured to obtain a second P-wave variance value corresponding to each lead at the current time according to the first P-wave variance value and a fourth parameter; and the eighteenth processing module is configured to obtain a fourth detection result according to the second P-wave variance value and the fourth average value, where the fourth detection result represents whether the cardiac beat type of the electrocardiographic data of each lead is a bundle branch block.

In an alternative embodiment, the decomposition module comprises: the device comprises a preprocessing module, a decomposition processing module and a determining module. The preprocessing module is used for preprocessing the electrocardio data to obtain preprocessed electrocardio data; the decomposition processing module is used for carrying out discrete wavelet decomposition on the preprocessed electrocardio data to obtain a plurality of wavelet coefficients; a determining module for determining at least one target wavelet coefficient from the plurality of wavelet coefficients.

In an alternative embodiment, the pre-processing module comprises: a nineteenth processing module, a twentieth processing module, and a twenty-first processing module. The nineteenth processing module is configured to perform first processing on the electrocardiographic data to obtain first electrocardiographic data, where the first processing at least includes one of: impedance matching processing, filtering processing and amplifying processing; a twentieth processing module, configured to perform second processing on the first electrocardiographic data to obtain second electrocardiographic data, where the second processing at least includes: analog-to-digital conversion processing; and the twenty-first processing module is used for carrying out low-pass filtering processing on the second electrocardiogram data to obtain preprocessed electrocardiogram data.

Example 4

According to another aspect of the embodiments of the present invention, there is also provided a storage medium including a stored program, wherein when the program runs, an apparatus in which the storage medium is controlled to execute the electrocardiographic signal detection method of embodiment 1 described above.

Example 5

According to another aspect of the embodiments of the present invention, there is further provided a processor, configured to execute a program, where the program executes the method for detecting an electrocardiographic signal according to embodiment 1.

The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.

In the above embodiments of the present invention, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.

In the embodiments provided in the present application, it should be understood that the disclosed technology can be implemented in other ways. The above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units may be a logical division, and in actual implementation, there may be another division, for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, units or modules, and may be in an electrical or other form.

The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.

In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.

The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a removable hard disk, a magnetic or optical disk, and other various media capable of storing program codes.

The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.

Claims

1. An electrocardiosignal detection method is characterized by comprising the following steps:

acquiring electrocardiogram data corresponding to a plurality of leads to obtain waveform characteristics corresponding to the electrocardiogram data, wherein the waveform characteristics comprise duration of QRS waves, probability density distribution of R waves and variability of P waves;

decomposing the electrocardio data, and determining at least one target wavelet coefficient according to the processing result;

respectively detecting the at least one target wavelet coefficient and the waveform characteristics based on a plurality of adaptive threshold detectors to obtain a plurality of detection results;

and performing heart beat type detection on the electrocardiogram data and the detection results based on a preset network model to obtain a target detection result, wherein the target detection result represents whether the heart beat type corresponding to the electrocardiogram data is a right bundle branch block.

2. The method of claim 1, wherein the plurality of adaptive threshold detectors comprises at least: a first detector based on wavelet coefficients, a second detector based on the waveform characteristics, a third detector based on the waveform characteristics, and a fourth detector based on the waveform characteristics, wherein the detection of the at least one target wavelet coefficient based on a plurality of adaptive threshold detectors results in a plurality of detection results, comprising:

detecting the at least one target wavelet coefficient based on the first detector to obtain a first detection result;

detecting the waveform characteristics based on the second detector to obtain a second detection result;

detecting the waveform characteristics based on the third detector to obtain a third detection result;

and detecting the waveform characteristics based on the fourth detector to obtain a fourth detection result.

3. The method of claim 2, wherein detecting the at least one target wavelet coefficient based on the first detector to obtain a first detection result comprises:

acquiring wavelet coefficient sums and wavelet coefficient products corresponding to each lead;

obtaining a first threshold corresponding to each lead according to the wavelet coefficient sum and the wavelet coefficient product;

averaging first threshold values corresponding to the plurality of leads to obtain a first average value;

obtaining a second threshold corresponding to each lead at the current moment according to the first threshold and the first parameter;

and obtaining the first detection result according to the second threshold and the first average value, wherein the first detection result represents whether the cardiac beat type of the electrocardiographic data of each lead is the right bundle branch block.

4. The method of claim 3, wherein obtaining a wavelet coefficient sum and a wavelet coefficient product corresponding to each lead comprises:

decomposing the electrocardiogram data to obtain a preset number of wavelet coefficients, wherein the preset number of wavelet coefficients at least comprises: a second wavelet coefficient, a third wavelet coefficient, a fourth wavelet coefficient, a fifth wavelet coefficient, a sixth wavelet coefficient, and a seventh wavelet coefficient, wherein the target wavelet coefficient at least comprises: the third wavelet coefficient, the fourth wavelet coefficient, the fifth wavelet coefficient;

summing the third wavelet coefficient, the fourth wavelet coefficient, the fifth wavelet coefficient and the sixth wavelet coefficient to obtain a wavelet coefficient sum;

summing the second wavelet coefficient, the sixth wavelet coefficient and the seventh wavelet coefficient to obtain a summation result;

and performing product operation on the third wavelet coefficient and the summation result to obtain the wavelet coefficient product.

5. The method of claim 2, wherein the waveform characteristics comprise at least: QRS wave duration, wherein detecting the waveform feature based on the second detector obtains a second detection result, including:

obtaining an S wave sequence and a Q wave sequence of waveform characteristics corresponding to each lead;

obtaining a first duration of a QRS wave corresponding to each lead according to the S wave sequence and the Q wave sequence;

averaging first duration corresponding to the plurality of leads to obtain a second average value;

obtaining a second duration corresponding to each lead according to the first duration and a second parameter;

and obtaining a second detection result according to the second duration and the second average value, wherein the second detection result represents whether the cardiac beat type of the electrocardiographic data of each lead is the right bundle branch block.

6. The method of claim 2, wherein the waveform characteristics comprise at least: an R-wave deviation value, wherein the detecting the waveform feature based on the third detector obtains a third detection result, including:

obtaining a first R wave deviation value corresponding to each lead according to the electrocardiogram data corresponding to each lead;

averaging first R wave skewness values corresponding to the plurality of leads to obtain a third average value;

obtaining a second R wave deviation value corresponding to each R sequence according to the first R wave deviation value and a third parameter;

and obtaining a third detection result according to the second R wave deviation value and the third average value, wherein the third detection result represents whether the cardiac beat type of the electrocardiographic data of each lead is the right bundle branch block.

7. The method of claim 6, wherein obtaining the first R-wave skewness value corresponding to each lead according to the electrocardiographic data corresponding to each lead comprises:

obtaining an R wave sequence in the electrocardiogram data corresponding to each lead;

calculating a kernel density estimation value vector corresponding to the R wave sequence;

and calculating the sequence length and the deviation value of the R wave sequence according to the kernel density estimated value vector.

8. The method of claim 2, wherein the waveform characteristics comprise at least: p-wave variance, wherein detecting the waveform feature based on the fourth detector to obtain a fourth detection result includes:

obtaining a first P wave variation value corresponding to each lead according to the electrocardiogram data corresponding to each lead;

averaging first P wave variation values corresponding to the plurality of leads to obtain a fourth average value;

obtaining a second P wave variation value corresponding to each lead at the current moment according to the first P wave variation value and a fourth parameter;

and obtaining a fourth detection result according to the second P wave variance value and the fourth average value, wherein the fourth detection result represents whether the cardiac beat type of the electrocardiographic data of each lead is the right bundle branch block.

9. An electrocardiosignal detection method is characterized by comprising the following steps:

acquiring electrocardiogram data corresponding to a plurality of leads, wherein waveform characteristics corresponding to the electrocardiogram data comprise duration of QRS waves, probability density distribution of R waves and variability of P waves;

respectively detecting the electrocardiogram data based on a plurality of self-adaptive threshold detectors to obtain a plurality of detection results;

10. An electrocardiographic signal detection device characterized by comprising:

the acquisition module is used for acquiring the electrocardiogram data corresponding to the leads to obtain waveform characteristics corresponding to the electrocardiogram data, wherein the waveform characteristics comprise duration of QRS waves, probability density distribution of R waves and variability of P waves;

the decomposition module is used for decomposing the electrocardio data and determining at least one target wavelet coefficient according to the processing result;

the first detection module is used for respectively detecting the at least one target wavelet coefficient and the waveform characteristics based on a plurality of adaptive threshold detectors to obtain a plurality of detection results;

and the second detection module is used for carrying out heart beat type detection on the electrocardio data and the detection results based on a preset network model to obtain a target detection result, wherein the target detection result represents whether the heart beat type corresponding to the electrocardio data is a right bundle branch block.

11. A storage medium, characterized in that the storage medium comprises a stored program, wherein when the program runs, the storage medium is controlled by a device to execute the electrocardiographic signal detection method according to any one of claims 1 to 8.

12. A processor, characterized in that the processor is configured to execute a program, wherein the program is configured to execute the method according to any one of claims 1 to 8 when running.