A method and system for identifying the degree of coronary stenosis by electro-mechanical coupling analysis

Abstract

The present application belongs to the technical field of physiological signal analysis, and proposes a method and system for recognizing the degree of coronary stenosis through electro-mechanical coupling analysis, which fully considers the performance of coronary stenosis on electrocardiogram and heart sound signals, extracts signal amplitude and interval sequences from the two kinds of signals, describes the activity state of the heart from the intensity and rhythm of the electrical and mechanical activities of the heart, and provides complementary information for recognizing coronary stenosis of different degrees; the multivariate transfer entropy of each amplitude sequence and interval sequence is calculated, the influence of coronary heart disease on electro-mechanical information transmission is quantified, and the comprehensiveness and accuracy of the evaluation of the degree of coronary stenosis in the later stage are improved.

Description

Technical Field

[0001] This invention belongs to the field of physiological signal analysis technology, and in particular relates to a method and system for identifying the degree of coronary artery stenosis by electro-mechanical coupling analysis. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] Coronary artery disease (CAD) is a major cause of rising global mortality. Long-term monitoring, early detection, accurate assessment, and appropriate treatment can effectively reduce CAD mortality. Coronary angiography can accurately measure the degree of coronary artery stenosis, but due to its invasiveness, high cost, and other drawbacks, it is not suitable for large-scale early screening of CAD.

[0004] Electrocardiogram (ECG) and heart sound signals reflect the physiological state of the heart from the perspectives of electrical and mechanical activity, respectively. When coronary artery stenosis occurs, these signals also undergo varying degrees of pathological changes. Due to the complex interactions within the human heart, there are a series of complex coupling relationships between the heart's two basic functions: electrical and mechanical. The electrical excitation of the heart can affect mechanical activities such as myocardial contraction through excitation-contraction coupling, while the mechanical activity of myocardial cells can affect the electrical excitation of myocardial cells and tissues through mechano-electrical feedback, thereby affecting the heart's electrical function.

[0005] Existing non-invasive methods are mostly based on extracting morphological features (such as heart rate variability) from single electrocardiogram or heart sound signals. Although there are also methods based on the joint analysis of electrocardiogram and heart sound signals, most of them only focus on the feature level and have limited exploration of the coupling relationship between electrocardiogram and heart sound. Summary of the Invention

[0006] To overcome the shortcomings of the prior art, this invention provides a method and system for identifying the degree of coronary artery stenosis through electromechanical coupling analysis. It fully considers the changes in amplitude intensity and interval rhythm of cardiac electrical and mechanical activities under coronary heart disease conditions, calculates multivariate transfer entropy, quantifies the impact of coronary heart disease on electromechanical information transmission, and improves the comprehensiveness and accuracy of subsequent assessment of the degree of coronary artery stenosis.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a method for identifying the degree of coronary artery stenosis using electromechanical coupling analysis, comprising: Multiple amplitude sequences and interval sequences were extracted from the synchronously acquired electrocardiogram and heart sound signals; Calculate the multivariate transition entropy of each amplitude sequence of ECG and heart sound signals, as well as the multivariate transition entropy of each interval sequence of ECG and heart sound signals, and construct an entropy feature sequence; Input the entropy feature sequence into the trained coronary artery stenosis identification model to obtain the coronary artery stenosis identification and classification results.

[0008] Secondly, the present invention provides a coronary artery stenosis degree identification system based on electromechanical coupling analysis, characterized in that it comprises: The extraction module is configured to extract multiple amplitude sequences and interval sequences from synchronously acquired electrocardiogram signals and heart sound signals, respectively. The calculation module is configured to: calculate the multivariate transfer entropy of each amplitude sequence of the electrocardiogram signal and the heart sound signal, as well as the multivariate transfer entropy of each interval sequence of the electrocardiogram signal and the heart sound signal, and construct the entropy value feature sequence; The recognition module is configured to input the entropy feature sequence into the trained coronary artery stenosis recognition model to obtain the coronary artery stenosis recognition classification result.

[0009] Thirdly, the present invention provides an electronic device including a memory and a processor, and computer instructions stored in the memory and running on the processor, wherein the computer instructions, when executed by the processor, perform the method described in the first aspect.

[0010] Fourthly, the present invention provides a computer-readable storage medium for storing computer instructions, which, when executed by a processor, perform the method described in the first aspect.

[0011] Fifthly, the present invention provides a computer program product, including a computer program that, when executed by a processor, implements the method described in the first aspect.

[0012] The above one or more technical solutions have the following beneficial effects: This invention fully considers the manifestations of coronary artery stenosis in both electrocardiogram (ECG) and heart sound signals. It extracts signal amplitude and interval sequences from these two signals, characterizing the heart's activity state from the perspectives of intensity and rhythm of cardiac electromechanical activity, providing complementary information for identifying different degrees of coronary artery stenosis. Furthermore, it calculates the multivariate transfer entropy of each amplitude and interval sequence, quantifying the impact of coronary artery disease on electromechanical information transmission, thus improving the comprehensiveness and accuracy of subsequent assessments of coronary artery stenosis severity.

[0013] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0014] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0015] Figure 1 This is a block diagram of the coronary artery stenosis identification system according to Embodiment 1 of the present invention; Figure 2 This is a flowchart illustrating the multivariate transfer entropy calculation in Embodiment 1 of the present invention; Detailed Implementation It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0016] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations of the present invention.

[0017] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0018] Terminology Explanation Multivariate transfer entropy is a causal nonlinear multivariate coupling method used to describe the nonlinear information transfer between multivariate time series, extending the transfer entropy metric originally proposed for paired systems to multiple interacting processes. Its aim is to identify all source nodes that have transferred effective information to the target node by recursively calculating conditional entropy.

[0019] Example 1 This embodiment discloses a method for identifying the degree of coronary artery stenosis using electromechanical coupling analysis, including: Acquire the synchronously acquired electrocardiogram and heart sound signals to be identified, and perform preprocessing; The amplitude and interval sequences of the preprocessed electrocardiogram and heart sound signals were extracted respectively. Multivariate transfer entropy is extracted based on amplitude and interval sequences, and entropy feature sequences and information transfer matrices are constructed. The entropy feature sequence is input into the trained coronary artery stenosis recognition model to obtain the coronary artery stenosis recognition result.

[0020] Based on the analysis of various amplitude and interval variability of cardiac cycle signals, this study reflects the cardiac state from the perspectives of both the intensity and rhythm of cardiac electromechanical activity. It provides crucial physiological information on cardiac systolic and diastolic function, neural regulation, and hemodynamic changes, thus offering strong support for early disease diagnosis and personalized treatment planning. Therefore, this embodiment fully considers the manifestations of coronary artery stenosis in both electrocardiogram and heart sound signals, as well as in both amplitude intensity and interval rhythm sequences. It calculates multivariate transfer entropy and extracts entropy features to quantitatively analyze the directionality and intensity of electromechanical interaction in the cardiovascular system, accurately assessing the degree of coronary artery stenosis.

[0021] Combination Figures 1-2 This embodiment discloses a method for identifying the degree of coronary artery stenosis through electromechanical coupling analysis, which includes the following steps: Step 1: Acquire the synchronously acquired ECG and heart sound signals to be identified and perform preprocessing; extract the amplitude sequence and interval sequence of the preprocessed ECG and heart sound signals respectively.

[0022] First, the electrocardiogram (ECG) and heart sound signals of the subjects were acquired simultaneously. A second-order Butterworth bandpass filter (0.05-75Hz) was used to remove noise interference from the ECG signal. A third-order polynomial Savitzky-Golay filter was used to process the baseline drift of the ECG signal. A notch filter (50Hz) was used to remove power frequency interference from the ECG signal. A Butterworth high-pass filter (20Hz) was used to remove low-frequency noise interference from the heart sound signal. A notch filter (50Hz) was used to remove power frequency interference from the heart sound signal.

[0023] For ECG signals, the start and peak points of the P wave, the start, end and peak points of the QRS wave, and the start, end and peak points of the T wave are located respectively. The amplitude sequences, including the P wave amplitude, R wave amplitude, T wave amplitude, ST segment amplitude sequence, and the interval sequences, including the RR interval, QT interval, PR interval and QRS interval sequence, are extracted.

[0024] For heart sound signals, locate the start and end points of the first heart sound (S1), systole, second heart sound (S2), and diastole in each cardiac cycle, extract the maximum amplitude of the S1 segment and the maximum amplitude of the S2 segment of the amplitude sequence, as well as the interval sequence including the S1 interval, S2 interval, systolic interval, and diastolic interval sequence.

[0025] Step 2: Calculate the multivariate transfer entropy of each amplitude sequence of the ECG and heart sound signals, as well as the multivariate transfer entropy of each interval sequence of the ECG and heart sound signals, and construct the entropy feature sequence and information transfer matrix.

[0026] Step 21: Combine the interval and amplitude sequences of the electrocardiogram (ECG) signal and the heart sound signal into a multivariate time series. To quantify the information transfer from the source node (e.g., a sequence X of the ECG signal) to the target node (e.g., a sequence Y of the heart sound signal), the influence of the other nodes Z needs to be controlled.

[0027] For example, the RR interval, PR interval, QT interval, etc. are extracted from the electrocardiogram signal, and the first heart sound interval, second heart sound interval, systolic interval, etc. are extracted from the heart sound signal. The information transmission from the RR interval (source node X) to the first heart sound interval (target node Y) is calculated, and the remaining sequence is Z.

[0028] When studying the effect of electrical activity on mechanical activity (electromechanical coupling): a sequence of electrocardiogram signals is usually selected as the source node X, and a sequence of heart sound signals is selected as the target node Y. This reflects the temporal process of cardiac electrical excitation triggering myocardial mechanical contraction.

[0029] When studying the effect of mechanical activity on electrical activity (mechanical-electrical feedback), the sequence of heart sound signals is selected as the source node X, and the sequence of electrocardiogram signals is selected as the target node Y. This reflects the reverse regulation of the mechanical tension changes of myocardial contraction on the electrical excitation of myocardial cells.

[0030] This method aims to calculate the additional information that the history of sequence X provides for predicting the current value of sequence Y, given the known history of node Z and the history of sequence Y itself; that is, the multivariate transition entropy. The history of node Z: A signal contains multiple cardiac cycles. The number of values ​​extracted from each sequence corresponds to the number of cardiac cycles. At the current time point t, the history consists of a series of values ​​before t.

[0031] Unlike traditional methods that construct lag candidate sets, this embodiment constructs a sample set. To prepare the data, among which For the source node's historical window, For the condition node historical window, The current value of the target node. k and This represents the length of the history window. The history window contains the first k values.

[0032] Regarding the interval sequences of electrocardiogram (ECG) and heart sound signals, the information transmission intensity from electrical activity to mechanical activity is higher, reflecting the timing and rhythm of cardiac contraction influenced by cardiac electrical signals, which in turn affects the speed and pressure changes of blood flow, ultimately affecting cardiac mechanical activity, corresponding to electromechanical coupling. Conversely, regarding the amplitude sequences of ECG and heart sound signals, the information transmission intensity from mechanical activity to electrical activity is higher, reflecting that the heart can influence the electrical potential excitation of myocardial cells and tissues through changes in the tension and amplitude of mechanical contraction, corresponding to mechanoelectric feedback.

[0033] Step 22: A rating-matching-based neural diffusion estimator is used to directly calculate the multivariate transition entropy. Its core is learning the probability distribution rating function of the target variable under specific conditions.

[0034] The scoring neural network is defined and trained as follows: Define a neural network Taking condition c as input, and the logarithmic gradient of the conditional probability density of the target variable y as the scoring function. The output is θ, where θ represents the network parameters.

[0035] For this task, two key conditions are defined: Condition A (Complete Condition): Let X represent the known history of the source node and condition node Z. The neural network outputs an estimate of the scoring function given the known history of the source node X and condition node Z.

[0036] Condition B (partial condition): , representing the condition node whose history is known only. The neural network outputs an estimate of the scoring function given only the history of condition node Z.

[0037] The neural network is trained using a noise reduction and scoring matching method. The training method is as follows: Data diffusion: for each training sample By adding noise to stochastic differential equations, the diffused variables are obtained. .

[0038] Loss function: Minimize the following denoising score matching loss:

[0039] in, It is a known noise distribution score, usually Gaussian.

[0040] The network is trained using a denoising score matching method, enabling it to accurately estimate under any condition c. conditional probability density The logarithmic gradient of the scoring function .

[0041] Step 23: Calculation of Transition Entropy: After training convergence, the multivariate transition entropy can be estimated using the following formula: ; Where ‖·‖² represents the square of the Euclidean norm, and N is the number of samples.

[0042] The amount of information transmitted from source node X to target node Y is proportional to the mean of the squared differences in the scoring function of Y under the two conditions of "having X history" and "not having X history". This difference directly measures the degree to which the introduction of X changes the description of the distribution of Y, i.e., the intensity of information transmission.

[0043] Step 24: Information Transfer Matrix Construction: Traverse all source and target node pairs to be analyzed (e.g., ECG interval → heart sound interval, ECG amplitude → heart sound amplitude, etc.), repeating steps 21-23 above to calculate the multivariate transfer entropy between each pair of nodes. Organize all calculation results according to node relationships to construct the information transfer matrix characterizing the intensity and direction of cardiac electromechanical information transmission, which will be used for subsequent coronary artery disease status identification.

[0044] The core of the method provided in this embodiment lies in training a scoring neural network. To directly approximate the probability distribution score function of the target variable Y under given condition c. The calculation of transfer entropy is transformed into the norm expectation of the difference between the output score vector of the neural network under two conditions: "including source node history" and "not including source node history". This process avoids direct estimation of the high-dimensional joint probability distribution, thus overcoming the "curse of dimensionality" problem faced by traditional methods when calculating the transfer entropy of multivariate signals such as ECG-heart sounds.

[0045] In summary, the multivariate transfer entropy calculated in this embodiment refers to the amount of information transferred from the source node X to the target node Y in a system with multiple time series nodes (such as different characteristic sequences of electrocardiograms and heart sounds), while controlling for the influence of other nodes Z. Unlike traditional methods that require constructing a lag candidate set and performing complex mutual information iteration calculations and statistical tests, this embodiment innovatively introduces a neural diffusion estimation method based on score matching. This method first constructs a sample set covering the historical and current values ​​of nodes, then trains a neural network to learn the conditional probability distribution characteristics of the system, and finally directly and efficiently quantifies the information transfer intensity through the differences in the forward propagation output of the neural network itself. This method is particularly suitable for information flow analysis of high-dimensional, nonlinear systems such as cardiac electromechanical coupling, and has higher estimation accuracy and computational efficiency. The multivariate transfer entropy calculation process is as follows: Figure 2 As shown.

[0046] Step 3: Input the entropy feature sequence into the trained coronary artery stenosis recognition model to obtain the coronary artery stenosis recognition and classification results.

[0047] In this embodiment, a feature recursive elimination method is used to select features from the entropy feature sequence. Based on the feature selection results, an optimal feature set is constructed and input into a one-to-many support vector machine OVA-SVM coronary artery stenosis identification model to achieve multi-level identification among different coronary artery stenosis groups.

[0048] Specifically, the OVA-SVM extends the binary support vector machine to multi-class tasks by training a separate binary classification model for each class, treating that class as the positive class and the other classes as the negative class.

[0049] During the testing phase, the input samples are classified by all binary classification models, and the category with the highest output classification score is selected as the final prediction result.

[0050] For the different coronary artery stenosis datasets constructed in this embodiment, OVA first treats samples of severe coronary artery stenosis (sCAD) as positive classes during training, and samples of moderate coronary artery stenosis (moCAD), mild coronary artery stenosis (miCAD), and healthy individuals as negative classes, constructing a training set for classification. Then, in turn, the moCAD, miCAD, and Health groups are treated as positive classes, and the rest as negative classes, constructing three more training sets and solving the corresponding SVM classifiers to ultimately achieve multi-class classification.

[0051] This embodiment fully considers the manifestations of coronary artery stenosis in both electrocardiogram (ECG) and heart sound signals, and combines their analysis to overcome the limitations of single signal analysis. By extracting both signal amplitude and interval sequences, it characterizes the cardiac activity state from the perspectives of intensity and rhythm of cardiac electromechanical activity, providing complementary information for the identification of different degrees of coronary artery stenosis. Multivariate transfer entropy is used to calculate the electromechanical information transmission matrix for different coronary artery disease groups, determining the intensity and direction of information transmission in the electromechanical network, and quantifying the impact of coronary artery disease on electromechanical information transmission, providing more comprehensive technical support for the early detection and accurate assessment of different degrees of coronary artery disease. Based on the traditional binary classification of disease and health, it introduces precise identification of different degrees of coronary artery stenosis, significantly improving the accuracy of diagnosis and its clinical application value.

[0052] Example 2 The purpose of this embodiment is to provide a method for identifying the degree of coronary artery stenosis using electromechanical coupling analysis, including: The extraction module is configured to extract multiple amplitude sequences and interval sequences from synchronously acquired electrocardiogram signals and heart sound signals, respectively. The calculation module is configured to: calculate the multivariate transfer entropy of each amplitude sequence of the electrocardiogram signal and the heart sound signal, as well as the multivariate transfer entropy of each interval sequence of the electrocardiogram signal and the heart sound signal, and construct the entropy value feature sequence; The recognition module is configured to input the entropy feature sequence into the trained coronary artery stenosis recognition model to obtain the coronary artery stenosis recognition classification result.

[0053] In further embodiments, the following is also provided: An electronic device includes a memory and a processor, as well as computer instructions stored in the memory and running on the processor. When executed by the processor, the computer instructions perform the method described in Embodiment 1. For brevity, further details are omitted here.

[0054] It should be understood that in this embodiment, the processor can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor, etc.

[0055] Memory may include read-only memory and random access memory, and provides instructions and data to the processor. A portion of memory may also include non-volatile random access memory. For example, memory may also store information about the device type.

[0056] A computer-readable storage medium for storing computer instructions, which, when executed by a processor, perform the method described in Embodiment 1.

[0057] The method in Embodiment 1 can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor. The software modules can reside in readily available storage media in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory; the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method. To avoid repetition, a detailed description is not provided here.

[0058] A computer program product includes a computer program that, when executed by a processor, implements the method described in Embodiment 1.

[0059] The present invention also provides at least one computer program product tangibly stored on a non-transitory computer-readable storage medium. The computer program product includes computer-executable instructions, such as instructions included in program modules, which execute in a device on a target real or virtual processor to perform the processes / methods described above. Typically, program modules include routines, programs, libraries, objects, classes, components, data structures, etc., that perform specific tasks or implement specific abstract data types. In various embodiments, the functionality of program modules can be combined or divided among program modules as needed. The machine-executable instructions for the program modules can execute within a local or distributed device. In a distributed device, the program modules can reside in both local and remote storage media.

[0060] The computer program code used to implement the methods of the present invention may be written in one or more programming languages. This computer program code may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the computer or other programmable data processing device, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a computer, partially on a computer, as a stand-alone software package, partially on a computer and partially on a remote computer, or entirely on a remote computer or server.

[0061] In the context of this invention, computer program code or related data may be carried by any suitable carrier to enable a device, apparatus, or processor to perform the various processes and operations described above. Examples of carriers include signals, computer-readable media, and the like. Examples of signals may include electrical, optical, radio, sound, or other forms of propagation signals, such as carrier waves, infrared signals, etc.

[0062] Those skilled in the art will recognize that the units and algorithm steps described in conjunction with the embodiments herein can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0063] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A method for identifying the degree of coronary artery stenosis using electromechanical coupling analysis, characterized in that, include: Multiple amplitude sequences and interval sequences were extracted from the synchronously acquired electrocardiogram and heart sound signals; Calculate the multivariate transition entropy of each amplitude sequence of ECG and heart sound signals, as well as the multivariate transition entropy of each interval sequence of ECG and heart sound signals, and construct an entropy feature sequence; Input the entropy feature sequence into the trained coronary artery stenosis identification model to obtain the coronary artery stenosis identification and classification results.

2. The method for identifying the degree of coronary artery stenosis by electromechanical coupling analysis as described in claim 1, characterized in that, The amplitude sequence of the electrocardiogram signal includes the amplitude sequences of the P wave, R wave, T wave, and ST segment; the interval sequence of the electrocardiogram signal includes the RR interval, QT interval, PR interval, and QRS interval; the amplitude sequence of the heart sound signal includes the maximum amplitude sequences of the S1 segment and the S2 segment; the interval sequence of the heart sound signal includes the S1 interval, S2 interval, systolic interval, and diastolic interval.

3. The method for identifying the degree of coronary artery stenosis by electromechanical coupling analysis as described in claim 1, characterized in that, The multivariate transition entropy of each amplitude sequence and interval sequence of the electrocardiogram signal is calculated as follows: For multivariate physiological time series, a training sample set is constructed, which includes the historical window of the source node, the current value of the target node, and the historical window of the condition node; The scoring neural network is trained based on the training sample set so that the scoring neural network learns the probability distribution scoring function of the target node under given conditions. By using a trained scoring neural network, the multivariate transfer entropy from the source node to the target node is estimated to quantify the information transmission strength between cardiac electrical activity and mechanical activity. Based on the calculated multiple transfer entropy values, an information transfer matrix characterizing the cardiac electromechanical information transfer network is constructed.

4. The method for identifying the degree of coronary artery stenosis by electromechanical coupling analysis as described in claim 3, characterized in that, The scoring neural network is trained under different given conditions, and the different given conditions include at least: the first condition is: including the source node history window and the condition node history window; The second condition is: it must only include the historical window of the condition node; The objective of training the scoring neural network is to make the scoring functions output by the scoring neural network approximate the logarithmic gradient of the conditional probability distribution of the target variable under the first and second conditions.

5. The method for identifying the degree of coronary artery stenosis by electromechanical coupling analysis as described in claim 1, characterized in that, Before extracting various amplitude sequences and interval sequences from the synchronously acquired ECG and heart sound signals, preprocessing of the synchronously acquired ECG and heart sound signals is also included. Specifically, a second-order Butterworth bandpass filter is used to remove noise interference from the ECG signal, a third-order polynomial Savitzky-Golay filter is used to process the baseline drift of the ECG signal, and a notch filter is used to remove power frequency interference from the ECG signal; a Butterworth high-pass filter is used to remove low-frequency noise interference from the heart sound signal, and a notch filter is used to remove power frequency interference from the heart sound signal.

6. The method for identifying the degree of coronary artery stenosis by electromechanical coupling analysis as described in claim 1, characterized in that, We used recursive feature elimination to select features from entropy feature sequences, and constructed a coronary artery stenosis identification model using a one-to-many support vector machine. Based on the coronary artery stenosis identification model, we evaluated the degree of coronary artery stenosis.

7. A coronary artery stenosis degree identification system based on electromechanical coupling analysis, characterized in that, include: The extraction module is configured to extract multiple amplitude sequences and interval sequences from synchronously acquired electrocardiogram signals and heart sound signals, respectively. The calculation module is configured to: calculate the multivariate transfer entropy of each amplitude sequence of the electrocardiogram signal and the heart sound signal, as well as the multivariate transfer entropy of each interval sequence of the electrocardiogram signal and the heart sound signal, and construct the entropy value feature sequence; The recognition module is configured to input the entropy feature sequence into the trained coronary artery stenosis recognition model to obtain the coronary artery stenosis recognition classification result.

8. An electronic device, characterized in that, It includes a memory and a processor, as well as computer instructions stored in the memory and running on the processor, which, when executed by the processor, perform the method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, Used to store computer instructions, which, when executed by a processor, perform the method described in any one of claims 1-6.

10. A computer program product, characterized in that, Includes a computer program, which, when executed by a processor, implements the method described in any one of claims 1-6.

Images