A classification model for myocardial ischemia based on 12-lead ECG, its construction method, and its application.

Abstract

This application proposes a classification model, construction method, and application of myocardial ischemia based on 12-lead ECG. By detecting the ST-T segment beat-by-beat changes using entropy domain, frequency domain, and Lyapunov domain analysis methods, spatiotemporal ECG feature parameters related to myocardial ischemia are obtained. A machine learning model for predicting myocardial ischemia is established, ultimately achieving optimal ECG feature selection related to myocardial ischemia. This addresses the problem that conventional 12-lead ECGs often lack clear ST segment or T wave features related to myocardial ischemia, severely affecting the sensitivity and accuracy of myocardial ischemia diagnosis. The method includes: acquiring a 12-lead ECG; calculating the ST-T segment sample entropy of the 12-lead ECG; converting the 12-lead ECG into a 3-lead ECG vector map; extracting the ST-T segment from the 3-lead ECG vector map and calculating its spatial and temporal feature values; and using a grid search method to select the optimal ECG features reflecting myocardial ischemia. Therefore, the sensitivity and accuracy of myocardial ischemia diagnosis are improved.

Description

Technical Field

[0001] This application relates to the field of electrocardiogram signal processing, and in particular to a myocardial ischemia classification model based on 12-lead ECG, its construction method, and its application. Background Technology

[0002] Myocardial ischemia is a relatively common cardiovascular disease that seriously threatens people's lives and health. It is a silent disease, meaning patients are unaware they have myocardial ischemia before it occurs, leading to a very high mortality rate. Therefore, early detection, early treatment, and early intervention are the most effective ways to reduce the mortality rate of myocardial ischemia.

[0003] Electrocardiography (ECG), as a surface record of cardiac electrical activity, contains rich physiological and pathological information. It is inexpensive, non-invasive, and convenient, making it the earliest and most fundamental method used in clinical detection of myocardial ischemia. However, the changes in the amplitude and phase of the T wave or ST segment on an ECG caused by myocardial ischemia during the heartbeat cycle are at the microvolt level and are not easily observed with the naked eye. Therefore, manual detection of myocardial ischemia based on 12-lead ECG has low sensitivity and accuracy.

[0004] Of course, some research reports have proposed using visual inspection methods to assist in the detection of electrocardiogram (ECG) abnormalities. Chinese patent CN201910364522.2 proposes an ECG abnormality assessment method based on ECG entropy maps. This method converts a 12-lead ECG into a three-dimensional ECG signal and then into an ECG entropy map to visually estimate the patient's ECG signal. However, this method does not clearly identify the type of ECG abnormality it detects; it can only provide a preliminary assessment and cannot solve the problem of accurately and sensitively identifying myocardial ischemia. In other words, there is currently no technical solution for accurately and sensitively identifying myocardial ischemia. Summary of the Invention

[0005] This application provides a myocardial ischemia classification model based on 12-lead ECG, its construction method, and its application. It constructs a myocardial ischemia classification model capable of early detection of myocardial ischemia. By inputting three features from ECG and VCG signals into the model, myocardial ischemia can be accurately classified. This solves the problem that conventional 12-lead ECG detection of myocardial ischemia lacks obvious ST segment or T wave features, severely affecting the sensitivity and accuracy of myocardial ischemia diagnosis.

[0006] In a first aspect, embodiments of this application provide a method for predicting myocardial ischemia based on 12-lead ECG, comprising the following steps:

[0007] Collect 12-lead ECG signals from the subjects and calculate the sample entropy of the ST-T segment in lead I of the ECG.

[0008] The 12-lead ECG signal is converted into a 3-lead VCG signal, and the temporal characteristic value TFV and spatial characteristic value SFV of the 3-lead VCG signal are calculated.

[0009] The sample entropy of the ST-T segment of ECG lead I, the temporal feature value (TFV), and the spatial feature value (SFV) are input into the trained myocardial ischemia classification model to obtain the output value. Based on the output value, it is predicted whether the subject has myocardial ischemia.

[0010] Secondly, embodiments of this application provide a method for constructing a myocardial ischemia classification model, comprising: establishing a training set: selecting the sample entropy of the ST-T segment of ECG lead I, the temporal feature value (TFV) of three-lead VCG, and the spatial feature value (SFV) of three-lead VCG from healthy individuals and patients with myocardial ischemia as the training set; inputting the training set into a support vector machine model for training, and obtaining the trained myocardial ischemia classification model, which can be used to predict whether myocardial ischemia has occurred.

[0011] Thirdly, embodiments of this application provide an application of a myocardial ischemia prediction method based on 12-lead ECG, including: a 12-lead electrocardiogram (ECG) device, an electronic data processing device, and a display component, wherein the electronic data processing device executes the myocardial ischemia prediction method based on 12-lead ECG.

[0012] Fourthly, embodiments of this application provide a myocardial ischemia prediction device based on a 12-lead ECG, comprising:

[0013] The ECG signal acquisition unit is used to acquire the subject's 12-lead ECG signal and calculate the sample entropy of the ST-T segment in lead I of the ECG.

[0014] The VCG signal acquisition unit is used to convert the 12-lead ECG signal into a 3-lead VCG signal and calculate the time characteristic value TFV and spatial characteristic value SFV of the 3-lead VCG signal.

[0015] The classification detection unit is used to input the sample entropy of the ST-T segment of the ECG lead I, the temporal feature value TFV, and the spatial feature value SFV into the trained myocardial ischemia classification model to obtain the output value, and predict whether the subject has myocardial ischemia based on the output value.

[0016] The main contributions and innovations of this invention are as follows:

[0017] This scheme uses entropy domain, frequency domain, and Lyapunov analysis methods to detect subtle changes in the ST-T segment induced by myocardial ischemia in both time and space, thus enabling early detection of myocardial ischemia. Temporal features include 12-lead ECG ST-T sample entropy, VCG ST-T sample entropy, and temporal feature TFV. Spatial features include VCG spatial feature SFV calculated using the Lyapunov index to detect spatial changes in VCG caused by myocardial ischemia, such as ST vector and T wave spatial changes. A combined ECG and VCG approach is used to extract time-space features related to myocardial ischemia from both ECG and VCG, thereby obtaining ECG spatiotemporal feature parameters associated with myocardial ischemia. The optimal ECG features for myocardial ischemia prediction are selected, and a machine learning model for myocardial ischemia prediction is established. During model construction, a grid search method is used to select the feature combination with the best classification performance, simplifying computation. The machine learning model for myocardial ischemia prediction and the corresponding myocardial ischemia prediction device built using this scheme can be used for early detection of myocardial ischemia, greatly facilitating medical diagnosis.

[0018] Details of one or more embodiments of this application are set forth in the following drawings and description to make other features, objects and advantages of this application more readily apparent. Attached Figure Description

[0019] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0020] Figure 1 This is a flowchart of a method for predicting myocardial ischemia based on 12-lead ECG according to an embodiment of this application;

[0021] Figure 2 This is a flowchart of the electrocardiogram signal processing method for medical devices used in the present invention for predicting myocardial ischemia based on 12-lead ECG;

[0022] Figure 3 The mean entropy of ST-T samples in different leads of healthy individuals and patients with myocardial ischemia in this embodiment of the invention;

[0023] Figure 4 This is a three-dimensional visualization image of the ST-T segment sequence of 3-lead VCG in an embodiment of the present invention;

[0024] Figure 5 This is a comparison chart of classification metrics between the ECG+VCG model and the ECG-only and VCG-only models in this embodiment of the invention.

[0025] Figure 6This is a structural block diagram of a myocardial ischemia prediction device based on 12-lead ECG according to an embodiment of this application;

[0026] Figure 7 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0027] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with one or more embodiments of this specification. Rather, they are merely examples of apparatuses and methods consistent with some aspects of one or more embodiments of this specification as detailed in the appended claims.

[0028] It should be noted that the steps of the corresponding methods are not necessarily performed in the order shown and described in this specification in other embodiments. In some other embodiments, the methods may include more or fewer steps than described in this specification. Furthermore, a single step described in this specification may be broken down into multiple steps in other embodiments; and multiple steps described in this specification may be combined into a single step in other embodiments.

[0029] Example 1

[0030] This application provides a method for predicting myocardial ischemia based on 12-lead ECG. Specifically, refer to... Figure 1 The method includes:

[0031] Collect 12-lead ECG signals from the subjects and calculate the sample entropy of the ST-T segment in lead I of the ECG.

[0032] The 12-lead ECG signal is converted into a 3-lead VCG signal, and the temporal characteristic value TFV and spatial characteristic value SFV of the 3-lead VCG signal are calculated.

[0033] The sample entropy of the ST-T segment of ECG lead I, the temporal feature value (TFV), and the spatial feature value (SFV) are input into the trained myocardial ischemia classification model to obtain the output value. Based on the output value, it is predicted whether the subject has myocardial ischemia.

[0034] This solution predicts myocardial ischemia from both temporal and spatial dimensions. It combines frequency analysis and sample entropy analysis to detect subtle changes in the ST-T segment over time induced by myocardial ischemia. It requires only three feature vectors to achieve accurate early detection of myocardial ischemia. It can be loaded into electrocardiographs for clinical applications or embedded in wearable devices for daily monitoring in nursing homes, health institutions, and homes. It has the advantages of high detection sensitivity and high detection accuracy.

[0035] In this scheme, conventional ECG acquisition equipment can be used to acquire 12-lead ECG signals from the subject. To facilitate subsequent ECG signal processing, in some embodiments, this scheme employs Butterworth high-pass filters and wavelet filters to perform noise reduction filtering on all ECG signals, removing power line interference, electromyographic interference, and baseline drift. In wavelet filtering, Coif4 from the Coifiet wavelet system is used as the wavelet basis, and the ECG signal is decomposed into four levels of Coif4. An adaptive threshold obtained using Stein's unbiased likelihood estimation principle is applied, followed by soft threshold filtering of the wavelet coefficients. Then, inverse wavelet transform is used to reconstruct the ECG signal, thus filtering out power line interference and electromyographic interference. All ECG signals are standardized according to 25 mm / s and 10 mm / mV standards.

[0036] In the step of "calculating the sample entropy of the ST-T segment of ECG lead I", the ST-T segment of lead I of the ECG signal is first extracted. The lead I ECG ST-T segment is then composed into a standard time series by heartbeat. Sub-vectors with an embedding dimension of m are extracted from the standard time series and formed into a first sub-vector sequence. The similarity probability of any sub-vector in the first sub-vector sequence with all other sub-vectors is calculated, and the average probability of all the sub-vectors is obtained by statistically analyzing the similarity probabilities of all the sub-vectors. The standard time series is then composed from the lead I ECG ST-T segment by heartbeat. Vectors with an embedding dimension of m+1 are extracted from the standard time series and formed into a second sub-vector sequence. The similarity probability of any sub-vector in the second sub-vector sequence with all other sub-vectors is calculated, and the average probability of all the sub-vectors is obtained by statistically analyzing the similarity probabilities of all the sub-vectors. The sample entropy of the lead I ST-T segment of the ECG is obtained based on the average probability of the first sub-vector sequence and the average probability of the second sub-vector sequence.

[0037] Specifically, in the steps of "forming a standard time series from the ST-T segment of lead I, extracting a sub-vector with an embedding dimension of m from the standard time series and forming a first sub-vector sequence" and "forming a standard time series from the ECG ST-T segment of lead I, extracting a vector with an embedding dimension of m+1 from the standard time series and forming a second sub-vector sequence", the first sub-vector sequence or the second sub-vector sequence can be extracted sequentially according to the arrangement order of the sub-vectors within the standard time series formed by the ST-T segment of lead I, based on the standard time series formed by the ST-T segment of lead I, based on the standard time series formed by the ST-T segment.

[0038] For example, if the standard time series composed of each ST-T segment beat in lead I is {x(j); 1≤j≤N}, where N is the sequence length of the standard time series composed of each ST-T segment beat in lead I:

[0039] At this point, from the ECG ST-T segment of lead I, a sub-vector with an embedding dimension of m is extracted from the standard time series composed of heartbeats to form the first sub-vector sequence X. m (1), ..., X m (N-m+1), where any subvector is: X m (i) = {x(i), x(i+1), ..., x(i+m-1)}, 1 ≤ i ≤ N-m+1, where N-m+1 represents the number of vectors and m represents the length of the extracted sequence.

[0040] Similarly, if we need to extract the second sub-vector sequence X with an embedding dimension of m+1... m+1 (1), ..., X m+1 (Nm), where any subvector is: X m+1 (i)={x(i), x(i+1),...,x(i+m)}, 1≤i≤Nm.

[0041] In the step of "calculating the similarity probability of any sub-vector in the first sub-vector sequence with all other sub-vectors", the distance between any sub-vector in the first sub-vector sequence and all other sub-vectors is calculated, the number of sub-vectors whose distance is less than a distance threshold is counted, and the product of the number of sub-vectors and the reciprocal of the number of all vectors is used as the similarity probability.

[0042] For example, compute any subvector X m (i) and any other subvector X m The distance to (j) is:

[0043]

[0044] Calculate any subvector X i and any subvector X j Similarity probability:

[0045]

[0046] In the formula, n i (m, T) is a vector X i and subvector X j The number of similar subvectors. Similarity is defined as d(X). i X j ) < T, where T is the distance threshold, which is the error tolerance range for similar regions.

[0047] Calculate the average probability:

[0048]

[0049] In the step of “obtaining the sample entropy of the ST-T segment of the ECG I lead based on the first average probability and the second average probability”, the negative value of the natural logarithm of the quotient of the second average probability and the first average probability is taken as the sample entropy of the ST-T segment of the ECG I lead.

[0050] For example,

[0051] Where m is the length of the sequence to be compared, corresponding to the sample embedding dimension; T is the distance threshold, corresponding to the error tolerance range of the similar region; and N is the sequence length of the standard time series of lead I.

[0052] In the step of "converting the 12-lead ECG signal into a 3-lead VCG signal", the following formula is used to convert the 12-lead ECG signal into a 3-lead VCG signal:

[0053]

[0054] Where I, II, V1, V2, V3, V4, V4, V5, and V6 represent the leads of the ECG signal, respectively; X, Y, and Z represent the leads of the VCG signal, respectively.

[0055] In the step of “calculating the spatial characteristic value SFV of the 3-lead VCG signal”, the VCGST-T segment of the 3-lead VCG signal is extracted and a three-dimensional time series is formed. The exponential rate of change of each data point in the three-dimensional time series is calculated, and the exponential rate of change is calculated based on the exponential rate of change.

[0056] Specifically, this scheme calculates the Lyapunov exponent for each point in the three-dimensional time series to obtain the exponential rate of change of the data at each point.

[0057] In the step of "calculating the exponential rate of change of each point in the three-dimensional time series", the distance between the nearest data point in the three-dimensional time series and the current data point is calculated to obtain an initial distance set; after increasing the number of steps for both the current data point and the corresponding nearest data point, an ending distance set is obtained; the logarithm of each data point in the initial distance set and the ending distance set is calculated to obtain the exponential rate of change of each data point.

[0058] In the step of "calculating the spatial feature value based on the exponential rate of change", all non-negative exponential rates of change for each data point are selected to obtain a set of non-negative exponential rates of change. The average value of the set of non-negative exponential rates of change is calculated to obtain the exponential average value for each data point. The exponential average value of all the data points is then taken to obtain the spatial feature value.

[0059] For example:

[0060] Extract the ST-T segment of the 3-lead VCG signal and form a three-dimensional time series: V i (t), i = Vx, Vy, Vz, the formula for calculating the exponential rate of change is as follows:

[0061] (1) Calculate the distance between the point in the three-dimensional sequence composed of VCG ST-T sequences that is spatially closest to the current k-th data point. This is the initial distance set.

[0062]

[0063] in, Indicates v k and The distance, v k This represents the current K data points. To be with v k The set of data points that are closest to it.

[0064] (2) Calculate the current data point v k and The distance after each step is increased by s steps; this is the set of final distances.

[0065]

[0066] in, Indicates v k+s and The distance, v k+s Used to represent the current v k The data points obtained after increasing the time by s steps, Used to indicate the current The data set obtained after increasing the time by s steps.

[0067] (3) Calculate the logarithm of each corresponding item in the initial distance set and the final distance set to obtain the exponential rate of change for each data point.

[0068]

[0069] (4) Select all non-negative exponential rates of change λ k Form a set Among them, z max This represents the number of all non-negative exponential rates of change.

[0070] (5) Calculate the average of the non-negative exponential rates of change.

[0071]

[0072] (6) Calculate the average to obtain the spatial eigenvalue (SFV).

[0073]

[0074] In the step of “calculating the time characteristic value TFV of the 3-lead VCG signal”, the Fourier amplitude spectrum of the ST-T segment sequence of the VCG signal of each lead is obtained, and the Fourier amplitude spectra of the three leads are integrated into the time characteristic value TFV.

[0075] In the process of "integrating the Fourier amplitude spectra of the three leads into a time feature value (TFV)," the Fourier amplitude spectrum of each lead is fitted to an exponential function, and then the time feature parameters are obtained by fitting the function. The time feature parameters of the three leads are then integrated to obtain the time feature value (TFV).

[0076] For example:

[0077] Extract the ST-T segment of the 3-lead VCG signal and form a three-dimensional time series: V i (t), i = Vx, Vy, Vz, the formula for calculating the time eigenvalue TFV is as follows:

[0078] (1) The ST-T segments of each lead in VCG form a one-dimensional array V i (t), i = Vx, Vy, Vz, perform Fourier transform on each one-dimensional signal and obtain the Fourier amplitude spectrum.

[0079] f i (w)=abs(F(X i ), i = 1, 2, 3;

[0080] (2) f i (w) The fit is an exponential function with λ as the exponent, thus obtaining the final characteristic parameter γ used for the fit. iUse it as a time feature parameter

[0081] (3) Calculate the time characteristic value TFV

[0082]

[0083] The myocardial ischemia classification model in this scheme adopts a support vector machine model, and the training method of the myocardial ischemia classification model is as follows:

[0084] Establish training set: Select the sample entropy of the ST-T segment of ECG lead I, the temporal characteristic value TFV of three-lead VCG, and the spatial characteristic value SFV of three-lead VCG from healthy individuals and patients with myocardial ischemia as training set;

[0085] Specifically, a training set is established, T = (x1, y1), ..., (x n y n ), y n ∈{1, -1}, where N represents the number of training samples, and the feature vector is defined as x n ;y n The values ​​∈{1, -1} correspond to the label data, where 1 represents that the sample belongs to myocardial ischemia, and -1 represents that the sample belongs to a healthy person. A linear regression function is established in the high-dimensional feature space, i.e., the formula y(x) = w is plotted. T The hyperplane depicted by x+b maximizes the boundaries between linear decision boundaries.

[0086] For each feature vector input into the support vector machine model, the support vector machine constructs a hyperplane to achieve linear segmentation for the binary classification problem.

[0087] w T .x n +β≥1 ify n =1

[0088] w T .x n +β<-1 ify n =-1

[0089] Where w represents the weight vector and β represents the offset.

[0090] To address the problem of linearly inseparable data, this invention employs the following constraints to establish the optimal separating hyperplane for the binary classification problem and a method for solving the quadratic optimization problem:

[0091]

[0092] The formula must satisfy:

[0093]

[0094]

[0095] Where C represents the hyperparameter penalty factor, and ξ is the slack variable.

[0096] Support Vector Machines transform the input space into a high-dimensional space, utilizing K(x) i x j )=φ(x i ) T φ(x j The kernel function is defined to create a separable hyperplane.

[0097] This invention selects the widely used Gaussian radial basis function as the kernel function, and its calculation method is as follows:

[0098]

[0099] Where σ is the parameter width.

[0100] Ultimately, the output of the support vector machine is the support vector with corresponding weight vectors, and the parameter deviation from the distance to the origin of the hyperplane.

[0101] This solution provides a myocardial ischemia classification model trained using the method described above.

[0102] Example 2

[0103] The selection of three dimensional vectors: ECG ST-T segment sample entropy, temporal feature value (TFV), and spatial feature value (SFV).

[0104] It is worth mentioning that the three dimensional vectors of ECG ST-T segment sample entropy, temporal feature value TFV, and spatial feature value SFV in this scheme are not arbitrarily chosen, but are obtained through model selection.

[0105] like Figure 2 As shown, the following describes the selection method for the three dimension vectors of ECG ST-T segment sample entropy, temporal feature value (TFV), and spatial feature value (SFV) in this scheme, including the following steps:

[0106] Collect 12-lead ECG signals from the subjects and calculate the sample entropy of the ECG ST-T segment of the 12-lead ECG signals;

[0107] The 12-lead ECG signal is converted into a 3-lead VCG signal, the ST-T segment of the 3-lead VCG signal is extracted, and the temporal characteristic value TFV and spatial characteristic value SFV of the VCG ST-T segment are calculated.

[0108] The 12-lead ECG ST-T segment was used as the input feature vector to train the support vector machine model. The model was used to classify patients with myocardial ischemia and healthy individuals. The grid search method was used to select the four ECG features that achieved the best classification performance.

[0109] The sample entropy, spatial feature value, and temporal feature value of the ST-T segment of the 3-lead ECG vector map are used as input feature vectors to train the support vector machine model. The model is used to classify patients with myocardial ischemia and healthy individuals. The grid search method is used to select the two VCG features that achieve the best classification performance.

[0110] The four ECG features and two VCG features are used as input feature vectors to train the support vector machine model, which is then used to classify patients with myocardial ischemia and healthy individuals. A grid search method is used to select the three features that achieve the best classification performance.

[0111] The methods for obtaining sample entropy, spatial feature values, and temporal feature values ​​have been described previously and will not be repeated here.

[0112] To differentiate between different support vector machine models, this scheme defines the support vector machine model that trains the 12-lead ECG ST-T segment as the ECG-only model, the support vector machine model that trains the 3-lead VCG ST-T segment as the VCG-only model, and the support vector machine model that trains both ECG and VCG simultaneously as the ECG+VCG model.

[0113] This scheme uses web search to obtain the optimal input vector for each model from all features, and compares the classification performance of the optimal input vector with that obtained by principal component analysis (PCA). After comparison, (SI, SII, SAVF, SV6) was found to have the best classification performance as input; therefore, this feature combination is the optimal input feature for the ECG-only model, as shown in Table 1. Similarly, (TFV, SFV) is the optimal input feature for the VCG-only model, as shown in Table 2. (SI, TFV, SFV) is the optimal input feature for the ECG+VCG-model, as shown in Table 3.

[0114] Table 1. Comparison of classification performance of different input vectors in the ECG-only model

[0115]

[0116] Table 2 compares the classification performance of different input vectors in the VCG-only model.

[0117]

[0118] Table 3 Comparison of classification performance with different input vectors in the ECG+VCG model

[0119]

[0120] The data obtained during the analysis process are shown in the following graphs. Figure 3-5 As shown, Figure 3 The difference is the mean entropy of 12-lead ECG ST-T samples between patients with myocardial ischemia and healthy individuals. It can be seen that the entropy of 12-lead ECG ST-T samples is higher in patients with myocardial ischemia than in healthy individuals. Figure 4 It is a three-dimensional visualization of the ST-T segment sequence of a 3-lead VCG. Figure 5 This is a comparison chart of classification metrics between the ECG+VCG model and the ECG-only and VCG-only models in embodiments of the present invention.

[0121] Example 3

[0122] This solution provides an application of a myocardial ischemia prediction method based on 12-lead ECG. In some embodiments, this method can be applied to a myocardial ischemia prediction device based on 12-lead ECG. In this case, the myocardial ischemia prediction device includes a 12-lead electrocardiogram (ECG) device, an electronic data processing device, and a display component. The electronic data processing device performs the following steps:

[0123] Acquire 12-lead ECG signals from the subject and calculate the sample entropy of the ST-T segment in lead I of the ECG.

[0124] The 12-lead ECG signal is converted into a 3-lead VCG signal, and the temporal characteristic value TFV and spatial characteristic value SFV of the 3-lead VCG signal are calculated.

[0125] The sample entropy of the ST-T segment of ECG lead I, the temporal feature value (TFV), and the spatial feature value (SFV) are input into the trained myocardial ischemia classification model to obtain the output value. Based on the output value, it is predicted whether the subject has myocardial ischemia.

[0126] The display component is configured to display a myocardial ischemia alarm or warning when it is determined that the 12-lead ECG data acquired by the 12-lead ECG device indicates myocardial ischemia.

[0127] The technical content of the method executed by the electronic data processing device is the same as that of Embodiment 1, and the repeated content will not be repeated here.

[0128] Example 4

[0129] Based on the same concept, referencing Figure 6This application also proposes a myocardial ischemia prediction device based on 12-lead ECG, comprising:

[0130] ECG signal acquisition unit 301 is used to acquire 12-lead ECG signals from the subject and calculate the sample entropy of the ST-T segment in ECG lead I.

[0131] VCG signal acquisition unit 302 is used to convert the 12-lead ECG signal into a 3-lead VCG signal and calculate the time characteristic value TFV and spatial characteristic value SFV of the 3-lead VCG signal.

[0132] The classification detection unit 303 is used to input the sample entropy of the ST-T segment of ECG lead I, the temporal feature value TFV and the spatial feature value SFV into the trained myocardial ischemia classification model to obtain the output value, and predict whether the subject has myocardial ischemia based on the output value.

[0133] The technical content in this embodiment four that is the same as that in embodiment one is described in the same way as in embodiment one, and the repeated content will not be repeated here.

[0134] Example 5

[0135] This embodiment also provides an electronic device, see reference. Figure 7 It includes a memory 404 and a processor 402, the memory 404 storing a computer program and the processor 402 being configured to run the computer program to perform the steps in any of the embodiments of the above-described method for predicting myocardial ischemia based on 12-lead ECG.

[0136] Specifically, the processor 402 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0137] The memory 404 may include a mass storage device for data or instructions. For example, and not limitingly, the memory 404 may include a hard disk drive (HDD), a floppy disk drive, a solid-state drive (SSD), flash memory, an optical disk drive, a magneto-optical disk drive, magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, the memory 404 may include removable or non-removable (or fixed) media. Where appropriate, the memory 404 may be internal or external to a data processing device. In a particular embodiment, the memory 404 is non-volatile memory. In a particular embodiment, the memory 404 includes read-only memory (ROM) and random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable read-only memory (PROM), an erasable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), an electrically alterable read-only memory (EAROM), or flash memory, or a combination of two or more of these. Where appropriate, the RAM can be Static Random-Access Memory (SRAM) or Dynamic Random-Access Memory (DRAM). DRAM can be Fast Page Mode Dynamic Random-Access Memory (FPMDRAM), Extended Data Out Dynamic Random-Access Memory (EDODRAM), Synchronous Dynamic Random-Access Memory (SDRAM), etc.

[0138] The memory 404 can be used to store or cache various data files that need to be processed and / or communicated, as well as possible computer program instructions executed by the processor 402.

[0139] The processor 402 reads and executes computer program instructions stored in the memory 404 to implement any of the myocardial ischemia prediction methods based on 12-lead ECG in the above embodiments.

[0140] Optionally, the electronic device may further include a transmission device 406 and an input / output device 408, wherein the transmission device 406 is connected to the processor 402, and the input / output device 408 is connected to the processor 402.

[0141] Transmission device 406 can be used to receive or send data via a network. Specific examples of the network described above may include wired or wireless networks provided by the communication provider of the electronic device. In one example, the transmission device includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, transmission device 406 may be a Radio Frequency (RF) module used for wireless communication with the Internet.

[0142] Input / output device 408 is used to input or output information. In this embodiment, the input information may be the acquired ECG signal and VCG signal, etc., and the output information may be the ST-T segment sample entropy, temporal feature value (TFV), spatial feature value (SFV), and myocardial ischemia classification results, etc.

[0143] Optionally, in this embodiment, the processor 402 can be configured to perform the following steps via a computer program:

[0144] S101. Collect the 12-lead ECG signal of the subject and calculate the sample entropy of the ST-T segment of ECG lead I;

[0145] S102. Convert the 12-lead ECG signal into a 3-lead VCG signal, and calculate the temporal characteristic value TFV and spatial characteristic value SFV of the 3-lead VCG signal.

[0146] S103. Input the sample entropy of the ST-T segment of the ECG lead I, the temporal feature value TFV, and the spatial feature value SFV into the trained myocardial ischemia classification model to obtain the output value, and predict whether the subject has myocardial ischemia based on the output value.

[0147] It should be noted that the specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementations, and will not be repeated here.

[0148] Generally, various embodiments can be implemented in hardware or dedicated circuitry, software, logic, or any combination thereof. Some aspects of the invention can be implemented in hardware, while others can be implemented in firmware or software that can be executed by a controller, microprocessor, or other computing device, but the invention is not limited thereto. Although various aspects of the invention may be shown and described as block diagrams, flowcharts, or using some other graphical representation, it should be understood that, by way of non-limiting example, these blocks, apparatuses, systems, techniques, or methods described herein can be implemented in hardware, software, firmware, dedicated circuitry or logic, general-purpose hardware or controllers or other computing devices, or some combination thereof.

[0149] Embodiments of the present invention can be implemented by computer software, which may be executable by a data processor of a mobile device, such as a processor entity, or by hardware, or by a combination of software and hardware. Computer software or programs (also referred to as program products), including software routines, applets, and / or macros, can be stored in any device-readable data storage medium, and they include program instructions for performing specific tasks. A computer program product may include one or more computer-executable components configured to perform embodiments when the program is run. One or more computer-executable components may be at least one piece of software code or a portion thereof. Additionally, it should be noted that any block in the logical flow of the figures may represent a program step, or interconnected logical circuitry, blocks and functions, or a combination of program steps and logical circuitry, blocks and functions. The software may be stored on physical media such as memory chips or blocks of storage implemented within a processor, magnetic media such as hard disks or floppy disks, and optical media such as, for example, DVDs and their data variants, CDs, etc. The physical medium is a non-transient medium.

[0150] Those skilled in the art should understand that the technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments have been described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0151] The above embodiments are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A myocardial ischemia prediction device based on 12-lead ECG, characterized in that, include: The ECG signal acquisition unit is used to acquire 12-lead ECG signals from the subject and calculate the sample entropy of the ST-T segment in lead I of the ECG. Specifically, the ST-T segment of lead I of the ECG signal is extracted, and a standard time series is formed by assembling the ST-T segment of lead I ECG heartbeat by heartbeat. Sub-vectors with an embedding dimension of m are extracted from the standard time series and formed into a first sub-vector sequence. The similarity probability of any sub-vector within the first sub-vector sequence is calculated against all other sub-vectors, and the average probability of all sub-vectors is calculated. A standard time series is formed from the ST-T segment of lead I ECG heartbeat by heartbeat. Vectors with an embedding dimension of m+1 are extracted from the standard time series and formed into a second sub-vector sequence. The similarity probability of any sub-vector within the second sub-vector sequence is calculated against all other sub-vectors, and the average probability of all sub-vectors is calculated. The sample entropy of the ST-T segment of lead I of the ECG is obtained based on the average probability of the first sub-vector sequence and the average probability of the second sub-vector sequence. The VCG signal acquisition unit is used to convert the 12-lead ECG signal into a 3-lead VCG signal, calculate the temporal characteristic value (TFV) and spatial characteristic value (SFV) of the 3-lead VCG signal, wherein the VCG ST-T segment of the 3-lead VCG signal is extracted and a three-dimensional time series is formed, the exponential rate of change of each data point in the three-dimensional time series is calculated, and the distance between the nearest data point in the three-dimensional time series and the current calculated data point is calculated to obtain an initial distance set; after increasing the number of steps for both the current calculated data point and the corresponding nearest data point, an ending distance set is obtained; the logarithm of each data point in the initial distance set and the ending distance set is calculated to obtain the exponential rate of change of each data point, all non-negative exponential rates of change for each data point are selected to obtain a non-negative exponential rate of change set, the average value of the non-negative exponential rate of change set is calculated to obtain the exponential average value of each data point, and the exponential average value of all the data points is taken to obtain the spatial characteristic value; The classification detection unit is used to input the sample entropy of the ST-T segment of the ECG lead I, the temporal feature value TFV, and the spatial feature value SFV into the trained myocardial ischemia classification model to obtain the output value, and predict whether the subject has myocardial ischemia based on the output value.

2. The myocardial ischemia prediction device based on 12-lead ECG according to claim 1, characterized in that, In the step of obtaining the sample entropy of the ST-T segment of ECG lead I based on the average probability of the first sub-vector sequence with embedding dimension m and the average probability of the second sub-vector sequence with embedding dimension m+1, the negative natural logarithm of the quotient of the average probability of the second sub-vector sequence and the average probability of the first sub-vector sequence is taken as the sample entropy of the ST-T segment of ECG lead I.

3. The myocardial ischemia prediction device based on 12-lead ECG according to claim 1, characterized in that, In the step of calculating the time characteristic value (TFV) of the 3-lead VCG signal, the Fourier amplitude spectrum of the ST-T segment sequence of the VCG signal in each lead is obtained, and the Fourier amplitude spectra of the three leads are integrated into the time characteristic value (TFV).

4. A method for constructing a classification model of myocardial ischemia, characterized in that, include: Establish a training set: Select the sample entropy of the ST-T segment of ECG lead I, the temporal feature value (TFV) of three-lead VCG, and the spatial feature value (SFV) of three-lead VCG from healthy individuals and patients with myocardial ischemia as the training set; input the training set into the support vector machine model for training to obtain a myocardial ischemia classification model. The trained myocardial ischemia classification model can be used to predict whether myocardial ischemia has occurred using the myocardial ischemia prediction device based on 12-lead ECG as described in any one of claims 1-3.

5. An application method for a myocardial ischemia prediction device based on 12-lead ECG, characterized in that, A myocardial ischemia prediction device based on 12-lead ECG, as described in any one of claims 1-3, collects 12-lead ECG signals from the subject and outputs whether the subject has myocardial ischemia.

Images