Coronary heart disease detection method and system based on cross-modal bidirectional coupling of electrocardio and heart sound signals

Abstract

The invention belongs to the technical field of signal analysis, and discloses an electrocardio and heart sound signal cross-modal bidirectional coupling coronary heart disease detection method and system, and the method comprises the steps: obtaining an electrocardio signal and a heart sound signal of a testee, carrying out the preprocessing, and extracting the time-frequency domain features of the obtained signals; the signal-to-noise ratio of the electrocardiosignal and the heart sound signal is calculated, the signal-to-noise ratio sensing weight is calculated according to the signal-to-noise ratio, cosine similarity calculation is carried out based on the signal-to-noise ratio sensing weight, and the instantaneous form similarity of the time domain features of the electrocardiosignal and the heart sound signal is quantified; calculating mutual information of the frequency domain characteristics of the electrocardiosignal and the heart sound signal, and generating a fusion weight according to the instantaneous form similarity of the mutual information and the time domain characteristics; and inputting the fusion weight into a classifier for coronary heart disease detection. By constructing the bidirectional cross-modal coupling model, a dynamic interaction mechanism between electrical activity and mechanical motion is disclosed, the limitation of traditional single-modal analysis is overcome, and the sensitivity and specificity of early diagnosis of the coronary heart disease are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of signal analysis technology, and in particular to a coronary heart disease detection method and system with cross-modal bidirectional coupling of electrocardiogram and heart sound signals. Background Art

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

[0003] Coronary artery disease (CAD) is a heart disease caused by vascular stenosis or obstruction due to coronary atherosclerosis, leading to myocardial ischemia and hypoxia. Clinical manifestations include angina pectoris, myocardial infarction, and even sudden death. The core pathological factor is the dynamic imbalance between abnormal electrical activity and impaired mechanical contractile function in cardiomyocytes. Traditional diagnostic methods rely primarily on clinical symptoms combined with imaging and electrophysiological studies, but these methods have significant limitations. While electrocardiograms (ECGs), which record myocardial electrical activity and detect ischemic features such as ST-segment depression and T-wave inversion, are noninvasive and convenient, their sensitivity under resting conditions is less than 50%. Coronary angiography, the gold standard for diagnosis, can directly visualize the location and extent of vascular stenosis. However, this procedure is invasive and carries the risk of puncture complications (e.g., hematoma, vascular injury) and contrast agent allergy. Its high cost makes it difficult to popularize in primary care settings or for early screening. Furthermore, these methods primarily focus on static assessment of structural stenosis and are unable to capture the dynamic changes in electromechanical coupling caused by myocardial ischemia in real time (e.g., the temporal correlation between heart murmurs and ECG abnormalities). This results in limited diagnostic capabilities for functional ischemia or microcirculatory disorders. Therefore, integrating electrophysiological and mechanical motion signals to make up for the singleness and lag of traditional methods has become an urgent problem to be solved for early warning and accurate classification of coronary heart disease. Summary of the Invention

[0004] In order to solve the above problems, the present invention proposes a coronary heart disease detection method and system with cross-modal bidirectional coupling of electrocardiogram and heart sound signals. By constructing a bidirectional cross-modal coupling model, the dynamic interaction mechanism between electrical activity and mechanical movement is revealed, overcoming the limitations of traditional single-modal analysis, and significantly improving the sensitivity and specificity of early diagnosis of coronary heart disease.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions: In a first aspect, the present invention provides a method for detecting coronary heart disease using bidirectional cross-modal coupling of electrocardiogram and heart sound signals, comprising the following steps: Synchronously acquiring the subject's electrocardiogram (ECG) signal and heart sound signal, preprocessing the acquired signals, and extracting the time-frequency domain features of the preprocessed ECG signal and heart sound signal respectively; Calculate the signal-to-noise ratio of the ECG signal and the heart sound signal respectively, calculate the signal-to-noise ratio perception weight according to the signal-to-noise ratio of the ECG signal and the heart sound signal, and perform cosine similarity calculation based on the signal-to-noise ratio perception weight to quantify the instantaneous morphological similarity of the time domain features of the ECG signal and the heart sound signal; Calculate the mutual information of the frequency domain features of the ECG signal and the heart sound signal, and generate the fusion weight based on the instantaneous morphological similarity of the mutual information and the time domain features; The fusion weights are input into the classifier for coronary heart disease detection.

[0006] As an optional implementation, preprocessing of the acquired ECG signal includes: first using adaptive morphological filtering, combined with a combination of opening and closing operations, to remove baseline drift and retain the QRS complex wave morphology, and then using the Teager energy operator to enhance the instantaneous energy characteristics of the R peak.

[0007] As an optional implementation, preprocessing the acquired heart sound signal includes: segmenting S1, S2 and murmur intervals based on state transition probability and observation probability through a hidden Markov model, and mapping the linear spectrum to a logarithmic Mel scale through a Mel filter bank.

[0008] As an optional implementation, the fusion weights are input into the classifier for coronary heart disease detection, specifically: A coupling strength threshold is established through support vector machine (SVM) training. When the fusion weight exceeds the threshold, it is determined that there is electro-mechanical coupling abnormality related to coronary heart disease.

[0009] As an optional implementation, when quantifying the instantaneous morphological similarity of the time domain features of the electrocardiogram signal and the heart sound signal, the electrocardiogram signal and the heart sound signal are dynamically time-warped and aligned.

[0010] As an optional embodiment, the cross-modal frequency band dependency is quantified by calculating the mutual information of the frequency domain features of the electrocardiogram signal and the heart sound signal through the joint probability distribution.

[0011] In a second aspect, the present invention provides a coronary heart disease detection system with cross-modal bidirectional coupling of electrocardiogram and heart sound signals, comprising: The data acquisition and preprocessing module is configured to: synchronously acquire the subject's electrocardiogram (ECG) signal and heart sound signal, preprocess the acquired signals, and extract the time-frequency domain features of the preprocessed ECG signal and heart sound signal; a similarity calculation module configured to: calculate the signal-to-noise ratios of the electrocardiogram signal and the heart sound signal respectively, calculate the signal-to-noise ratio perception weights according to the signal-to-noise ratios of the electrocardiogram signal and the heart sound signal, and perform cosine similarity calculation based on the signal-to-noise ratio perception weights to quantify the instantaneous morphological similarity of the time domain features of the electrocardiogram signal and the heart sound signal; The fusion weight generation module is configured to: calculate the mutual information of the frequency domain features of the electrocardiogram signal and the heart sound signal, and generate the fusion weight according to the mutual information and the instantaneous morphological similarity of the time domain features; The diagnosis decision module is configured to input the fusion weight into the classifier for coronary heart disease detection.

[0012] In a third aspect, the present invention provides an electronic device comprising a memory and a processor, and computer instructions stored in the memory and executed on the processor, wherein the computer instructions, when executed by the processor, perform the method described in the first aspect.

[0013] In a fourth aspect, the present invention provides a computer-readable storage medium for storing computer instructions, wherein when the computer instructions are executed by a processor, the method described in the first aspect is performed.

[0014] In a fifth aspect, the present invention provides a computer program product, comprising a computer program, which implements the method described in the first aspect when executed by a processor.

[0015] Compared with the prior art, the present invention has the following beneficial effects: The present invention proposes a coronary heart disease detection method and system with cross-modal bidirectional coupling of ECG and heart sound signals. It fully considers the manifestation of coronary heart disease in both ECG and heart sound signals, and conducts a joint analysis of the two for multimodal fusion, aiming to break through the limitations of traditional single-modal analysis, synchronously integrate ECG electrical activity and PCG mechanical motion signals, and accurately capture the "electro-mechanical" causal closed loop, including forward ECG→PCG drive and reverse PCG→ECG feedback. In terms of dynamic modeling, the present invention uses DTW to solve the asynchrony of electro-mechanical conduction, and combines mutual information to mine implicit frequency domain correlations, thereby achieving real-time tracking of the dynamic process of myocardial ischemia. At the same time, a dynamic weight allocation mechanism based on signal-to-noise ratio is used for adaptive weight adjustment to improve the reliability of feature fusion in different noise environments, and adapt to portable devices in primary medical scenarios. The present invention quantifies the coupling strength through a multi-scale similarity function for early diagnosis, which can significantly improve the sensitivity of coronary heart disease diagnosis in the resting state.

[0016] Advantages of additional aspects of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0018] Figure 1This is a flow chart of the coronary heart disease detection method using cross-modal bidirectional coupling of electrocardiogram and heart sound signals provided in Example 1 of the present invention; Figure 2 This is a flowchart for constructing a bidirectional cross-modal coupling model provided in Example 1 of the present invention. DETAILED DESCRIPTION

[0019] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0020] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0021] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but includes other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0022] In the absence of conflict, the embodiments of the present invention and the features thereof may be combined with each other.

[0023] Example 1 like Figure 1 As shown, this embodiment provides a coronary heart disease detection method with cross-modal bidirectional coupling of electrocardiogram and heart sound signals, comprising the following steps: Synchronously acquiring the subject's electrocardiogram (ECG) signal and heart sound signal, preprocessing the acquired signals, and extracting the time-frequency domain features of the preprocessed ECG signal and heart sound signal respectively; Calculate the signal-to-noise ratio of the ECG signal and the heart sound signal respectively, calculate the signal-to-noise ratio perception weight according to the signal-to-noise ratio of the ECG signal and the heart sound signal, and perform cosine similarity calculation based on the signal-to-noise ratio perception weight to quantify the instantaneous morphological similarity of the time domain features of the ECG signal and the heart sound signal; Calculate the mutual information of the frequency domain features of the ECG signal and the heart sound signal, and generate the fusion weight based on the instantaneous morphological similarity of the mutual information and the time domain features; The fusion weights are input into the classifier for coronary heart disease detection.

[0024] When coronary artery disease causes myocardial ischemia, the correlation between cardiac electrical activity (ECG) and mechanical motion (PCG) becomes increasingly pronounced. Coronary artery stenosis leads to insufficient myocardial blood supply, impairing ion channel function in the ischemic area, slowing electrical conduction (e.g., sodium-potassium pump dysfunction). ECG findings include widened QRS complexes (delayed ventricular depolarization) or ST-T segment abnormalities (repolarization disturbances). This electrical conduction delay further leads to spatiotemporal asynchrony in ventricular contraction: Parts of the myocardium contract late due to electrical signal lags, while adjacent areas have already entered diastole. This mechanical asynchrony results in abnormal ventricular wall tension, which can be detected by PCG as a systolic murmur (due to turbulent blood flow) or splitting of heart sounds (asynchrony between left and right ventricular contractions). For example, left bundle branch block (ECG QRS complex ≥120ms) is often accompanied by a weakened and paradoxical splitting of the S1 sound (aortic valve closure later than the pulmonary valve) on PCG.

[0025] The bidirectional cross-modal mechanism is a computational framework that simultaneously models the bidirectional interaction between two different modal data (such as electrocardiogram (ECG) and heart sound (PCG). Its core is to capture the causal loop between electrical and mechanical activity: Forward path (ECG→PCG): Analyzes how electrical signals drive mechanical activity (e.g., heart contraction triggering heart sounds).

[0026] Reverse pathway (PCG→ECG): exploring how mechanical activity can feedback and affect electrical signals (e.g., abnormal blood flow leading to changes in ECG morphology).

[0027] Given two modal signal spaces X (e.g., ECG) and Y (e.g., PCG), the bidirectional cross-modal mechanism aims to construct bidirectional mapping functions f(X)→Y, g(Y)→X, satisfying causal constraints (e.g., the input-output relationship between f and g must conform to the physiological causal chain) and dynamic interactivity (the bidirectional mapping needs to model time-varying dependencies). By jointly modeling bidirectional relationships, the dynamic dependencies of physiological processes can be fully revealed.

[0028] The technical solution of the present invention is described in detail below: First, the electrocardiogram (ECG) signal and the heart sound signal of the test subject are acquired, and the acquired signals are preprocessed.

[0029] Among them, the preprocessing operations for electrocardiogram (ECG) signals include adaptive morphological filtering, combining opening operations (erosion followed by dilation) and closing operations (dilation followed by erosion) to remove baseline drift, preserve the QRS complex morphology, and use wavelet threshold denoising (sym4 wavelet, 5-layer decomposition) to eliminate myoelectric interference. Feature detection is used for ECG, integrating the Teager Energy Operator (TEO) and the Pan-Tompkins algorithm to improve detection accuracy under low signal-to-noise ratios and enhance the instantaneous energy characteristics of the R peak. The formula is:

[0030] Where: x[n] is the ECG signal.

[0031] Perform R peak detection on ECG, segment the P-QRS-T complex wave according to the R peak position, and extract the time domain features of each band (such as ST segment slope and QT interval). Finally, output the R peak time series and time domain features (such as QRS width Δt QRS , ST segment slope k ST ).

[0032] The preprocessing operation of the heart sound signal includes segmenting and filtering the PCG. The heart sound signal is first substituted into the hidden Markov model (HMM) and Mel spectrogram processing. The hidden Markov model (HMM) is used to segment S1, S2, systole, and diastole. The Mel filter group (20-600Hz) is used to separate the pathological related frequency bands (such as the high-frequency components of aortic stenosis), segment the heart sound components and extract acoustic features. Among them, the hidden Markov model is used to extract the acoustic features based on the state transition probability. and the observation probability , split S1, S2 and noise interval; through the Mel filter bank {H m (f)} Map the linear spectrum to the logarithmic Mel scale, the formula is:

[0033] Where X(f) is the PCG spectrum, H m (f) is the frequency domain feature extracted by Mel filter.

[0034] Finally output S1 / S2 start time {t S1 , t S2} and MFCC feature vector.

[0035] Based on the preprocessed data, the bidirectional interaction between ECG and PCG is modeled using a multi-scale similarity calculation function (sim function). The specific steps include: Step 1: Perform multi-scale feature mapping.

[0036] On the time domain feature map: extract the ECG time domain feature vector Q t EGG , Q t EGG =[R peak time, QRS width, ST slope, T wave amplitude] T , extract the PCG time domain feature vector K t PCG , K t PCG =[S1 onset time, diastolic duration, murmur duration] T。

[0037] In the frequency domain feature mapping, the ECG frequency domain is decomposed by wavelet algorithm, and the Mel spectrum (focusing on the pathological frequency band) is calculated to decompose the PCG frequency domain.

[0038] Step 2: Perform multi-scale similarity calculation.

[0039] Calculate the signal-to-noise ratio (SNR) of the ECG and heart sound signals respectively, and improve the reliability of the dynamic balance time domain and frequency domain features through the perception weight. ECG ), PCG signal-to-noise ratio (SNR PCG ), calculate the signal-to-noise ratio perception weight: , .

[0040] Cosine similarity calculation is performed to match the ECG and PCG morphology, and the instantaneous morphological similarity of the time domain feature vectors of ECG and PCG is quantified. The specific formula is:

[0041] Among them, Q i , K j are ECG and PCG time domain feature vectors, β1 and β2 are ECG and PCG weights respectively.

[0042] Furthermore, before quantifying the instantaneous morphological similarity of time domain features, it is necessary to perform dynamic time warping (DTW) alignment on the electrocardiogram (ECG) and cardiac sound (PCG) signals. This step is used to solve the asynchrony problem in the electromechanical conduction process by defining the cost matrix , and uses a dynamic programming algorithm to find the optimal path, constraining the path offset to not exceed the physiological delay window (such as 50-200 ms), and aligning the signals on the time axis. This ensures that subsequent similarity calculations are performed between time-aligned signals, thereby improving the accuracy of the analysis.

[0043] Furthermore, by calculating the frequency domain mutual information (MI), through the joint probability distribution p(x,y), we can quantify the cross-modal frequency band dependency, such as the statistical dependency between ECG low-frequency fluctuations (such as ST segment changes of 0.5-5 Hz) and PCG high-frequency murmurs (such as 300-600 Hz), capture hidden pathological associations, and avoid the limitations of linear correlation. The specific formula is:

[0044] Among them, Q f is the frequency domain feature set of ECG, K f is the frequency domain feature set of PCG, p(x,y) is the joint probability distribution (x∈Q f ,y∈K fThe probability of simultaneous occurrence), p(x)p(y) is the product of the marginal probability distributions.

[0045] Step 3: Dynamic weight allocation and fusion.

[0046] The model structure achieves linear time series alignment by calculating the CosSim branch, and completes nonlinear frequency domain statistics by calculating MI, realizing complementary enhancement of cross-modal features.

[0047] The values ​​of CosSim and MI are normalized to [0, 1] to eliminate dimensionality differences. Fusion weights are generated based on the modal signal-to-noise ratio of the current sample.

[0048] The weighted fusion function is:

[0049] in: , .

[0050] When high SNR ECG With low SNR PCG When the pathology is low (e.g., myocardial ischemia, due to the nonlinear relationship between ST segment changes and murmurs), the MI weight is increased, focusing on the ECG-PCG frequency domain correlation.

[0051] Finally, Sim fusion The results are directly input into the classifier SVM as a diagnostic indicator to achieve cross-modal joint analysis. fusion >Specific thresholds can be used to determine "pathological association." This can assist clinicians in early warning and accurate classification of coronary heart disease.

[0052] Example 2 This embodiment provides a coronary heart disease detection system with cross-modal bidirectional coupling of electrocardiogram and heart sound signals, including: The data acquisition and preprocessing module is configured to: synchronously acquire the subject's electrocardiogram (ECG) signal and heart sound signal, preprocess the acquired signals, and extract the time-frequency domain features of the preprocessed ECG signal and heart sound signal; a similarity calculation module configured to: calculate the signal-to-noise ratios of the electrocardiogram signal and the heart sound signal respectively, calculate the signal-to-noise ratio perception weights according to the signal-to-noise ratios of the electrocardiogram signal and the heart sound signal, and perform cosine similarity calculation based on the signal-to-noise ratio perception weights to quantify the instantaneous morphological similarity of the time domain features of the electrocardiogram signal and the heart sound signal; The fusion weight generation module is configured to: calculate the mutual information of the frequency domain features of the electrocardiogram signal and the heart sound signal, and generate the fusion weight according to the mutual information and the instantaneous morphological similarity of the time domain features; The diagnosis decision module is configured to input the fusion weight into the classifier for coronary heart disease detection.

[0053] It should be noted that the above modules correspond to the steps described in Example 1, and the examples and application scenarios implemented by the above modules and the corresponding steps are the same, but are not limited to the contents disclosed in the above Example 1. It should be noted that the above modules, as part of the system, can be executed in a computer system such as a set of computer-executable instructions.

[0054] In further embodiments, there is also provided: An electronic device includes a memory and a processor, and computer instructions stored in the memory and executed by the processor, wherein when the computer instructions are executed by the processor, the method described in Example 1 is performed. For the sake of brevity, no further details are given here.

[0055] It should be understood that in this embodiment, the processor may be a central processing unit (CPU), or may be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), off-the-shelf field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0056] The memory may include a read-only memory and a random access memory, and provides instructions and data to the processor. A portion of the memory may also include a non-volatile random access memory. For example, the memory may also store information about the device type.

[0057] A computer-readable storage medium is used to store computer instructions, and when the computer instructions are executed by a processor, the method described in Example 1 is performed.

[0058] The method in Example 1 can be directly implemented as a hardware processor, or can be implemented using a combination of hardware and software modules in the processor. The software module can be located in a storage medium mature in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, etc. The storage medium is located in the memory, and the processor reads the information in the memory and, in conjunction with its hardware, completes the steps of the above method. To avoid repetition, it will not be described in detail here.

[0059] A computer program product includes a computer program, which implements the method described in embodiment 1 when executed by a processor.

[0060] 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 contained in program modules, which are executed in a device on a real or virtual processor of a target to perform the process / method described above. Generally, 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 between program modules as needed. The machine-executable instructions for the program modules can be executed in local or distributed devices. In distributed devices, program modules can be located in local and remote storage media.

[0061] The computer program code for implementing the method of the present invention can be written in one or more programming languages. These computer program codes can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device so that when the program code is executed by the computer or other programmable data processing device, the functions / operations specified in the flow chart and / or block diagram are implemented. The program code can be executed entirely on a computer, partially on a computer, as an independent software package, partially on a computer and partially on a remote computer, or entirely on a remote computer or server.

[0062] In the context of the present invention, computer program code or related data can be carried by any appropriate 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 include electrical, optical, radio, acoustic, or other forms of propagated signals, such as carrier waves, infrared signals, and the like.

[0063] Those skilled in the art will appreciate that the units and algorithm steps of the various examples described in conjunction with this embodiment can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians 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.

[0064] Although the above describes the specific embodiments of the present invention in conjunction with the accompanying drawings, it 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 on the basis of the technical solution of the present invention without any creative work are still within the scope of protection of the present invention.

Claims

1. A coronary heart disease detection method using bidirectional cross-modal coupling of electrocardiogram and heart sound signals, characterized in that: The following steps are involved: Synchronously acquiring the subject's electrocardiogram (ECG) signal and heart sound signal, preprocessing the acquired signals, and extracting the time-frequency domain features of the preprocessed ECG signal and heart sound signal respectively; Calculate the signal-to-noise ratio of the ECG signal and the heart sound signal respectively, calculate the signal-to-noise ratio perception weight according to the signal-to-noise ratio of the ECG signal and the heart sound signal, and perform cosine similarity calculation based on the signal-to-noise ratio perception weight to quantify the instantaneous morphological similarity of the time domain features of the ECG signal and the heart sound signal; Calculate the mutual information of the frequency domain features of the ECG signal and the heart sound signal, and generate the fusion weight based on the instantaneous morphological similarity of the mutual information and the time domain features; The fusion weights are input into the classifier for coronary heart disease detection.

2. The coronary heart disease detection method using cross-modal bidirectional coupling of electrocardiogram and heart sound signals according to claim 1, wherein: The preprocessing of the acquired ECG signal includes: firstly using adaptive morphological filtering, combining opening and closing operations to remove baseline drift and retain the QRS complex wave morphology, and then using the Teager energy operator to enhance the instantaneous energy characteristics of the R peak.

3. The coronary heart disease detection method using cross-modal bidirectional coupling of electrocardiogram and heart sound signals according to claim 1, wherein: The preprocessing of the acquired heart sound signal includes: segmenting the S1, S2 and murmur intervals based on the state transition probability and observation probability through the hidden Markov model, and mapping the linear spectrum to the logarithmic Mel scale through the Mel filter bank.

4. The coronary heart disease detection method using cross-modal bidirectional coupling of electrocardiogram and heart sound signals according to claim 1, wherein: The fusion weights are input into the classifier for coronary heart disease detection, specifically: A coupling strength threshold is established through support vector machine (SVM) training. When the fusion weight exceeds the threshold, it is determined that there is electro-mechanical coupling abnormality related to coronary heart disease.

5. The coronary heart disease detection method using cross-modal bidirectional coupling of electrocardiogram and heart sound signals according to claim 1, wherein: When quantifying the instantaneous morphological similarity of the time domain features of the ECG signal and the heart sound signal, the ECG signal and the heart sound signal are dynamically time-warped and aligned.

6. The coronary heart disease detection method using cross-modal bidirectional coupling of electrocardiogram and heart sound signals according to claim 1, wherein: By calculating the mutual information of the frequency domain features of the ECG signal and the heart sound signal, the cross-modal frequency band dependency is quantified through the joint probability distribution.

7. A coronary heart disease detection system with cross-modal bidirectional coupling of electrocardiogram and heart sound signals, characterized in that: include: The data acquisition and preprocessing module is configured to: synchronously acquire the subject's electrocardiogram (ECG) signal and heart sound signal, preprocess the acquired signals, and extract the time-frequency domain features of the preprocessed ECG signal and heart sound signal; a similarity calculation module configured to: calculate the signal-to-noise ratios of the electrocardiogram signal and the heart sound signal respectively, calculate the signal-to-noise ratio perception weights according to the signal-to-noise ratios of the electrocardiogram signal and the heart sound signal, and perform cosine similarity calculation based on the signal-to-noise ratio perception weights to quantify the instantaneous morphological similarity of the time domain features of the electrocardiogram signal and the heart sound signal; The fusion weight generation module is configured to: calculate the mutual information of the frequency domain features of the electrocardiogram signal and the heart sound signal, and generate the fusion weight according to the mutual information and the instantaneous morphological similarity of the time domain features; The diagnosis decision module is configured to input the fusion weight into the classifier for coronary heart disease detection.

8. An electronic device, characterized in that: The method comprises a memory and a processor, and computer instructions stored in the memory and executed on the processor, wherein when the computer instructions are executed by the processor, the method according to any one of claims 1 to 6 is completed.

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

10. A computer program product, characterized in that The invention comprises a computer program, which is used to implement the method according to any one of claims 1 to 6 when the computer program is executed by a processor.

Images