Electrocardiosignal generation method, cardiovascular disease diagnosis apparatus, device, and medium

Abstract

The application discloses an electrocardiosignal generation method, a cardiovascular disease diagnosis device, equipment and a medium. The method can generate multi-lead electrocardiosignal based on electrocardiosignal collected in a lead mode, does not need a user to wear a complex patch for electrocardiosignal collection, is suitable for an application scene of a smart wearable device, can obtain relatively comprehensive electrocardiosignal, and is better in practicability. Moreover, the method can evaluate and select an optimal lead mode suitable for serving as a signal generation basis, can improve the accuracy of generated electrocardiosignal, and facilitates efficient and reliable monitoring of a health state of a user based on the generated electrocardiosignal. The application can be widely applied in the field of artificial intelligence technology.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a method for generating electrocardiogram signals, a device, equipment, and medium for diagnosing cardiovascular diseases. Background Technology

[0002] Cardiovascular disease is one of the leading causes of death worldwide. Electrocardiography (ECG) is commonly used for the early diagnosis of cardiovascular disease and plays a crucial role in reducing patient mortality. Healthcare professionals can diagnose cardiovascular disease by analyzing typical changes in ECGs, such as abnormalities in characteristic waves like the P wave, QRS complex, and ST segment.

[0003] Currently, electrocardiograms (ECGs) commonly used in medicine include multi-lead ECG signals, which can comprehensively reflect the health status of the heart and help doctors make accurate clinical diagnoses. However, acquiring standard multi-lead ECG signals is difficult; for example, acquiring 12-lead ECG signals requires numerous patches, restricting the patient's freedom of movement. This is only suitable for clinical medical diagnosis and is not conducive to long-term cardiovascular disease monitoring. To achieve long-term monitoring of cardiovascular diseases, everyday wearable ECG signal measurement devices (such as smart bracelets and patches) have been developed and used. These everyday wearable devices can monitor ECG signals in real time while minimizing the impact on patients' normal lives. However, everyday wearable ECG signal measurement devices have a significant limitation: most of them can only measure ECG signals from a single lead, resulting in lower practicality and difficulty in comprehensively reflecting the health status of the heart.

[0004] In summary, the problems with the relevant technologies urgently need to be addressed. Summary of the Invention

[0005] The purpose of this application is to at least partially solve one of the technical problems existing in the related art.

[0006] Therefore, one object of the embodiments of this application is to provide a method for generating electrocardiogram signals, a device for diagnosing cardiovascular diseases, an apparatus, and a medium.

[0007] To achieve the above-mentioned technical objectives, the technical solutions adopted in the embodiments of this application include:

[0008] On one hand, embodiments of this application provide a method for generating electrocardiogram (ECG) signals, the method comprising:

[0009] Acquire multi-lead electrocardiogram (ECG) signals corresponding to multiple sample individuals; wherein, the multi-lead ECG signals include the first ECG signals measured in various lead modes, and there are multiple types of lead modes;

[0010] The first electrocardiogram (ECG) signal corresponding to the first lead mode is input into the ECG signal generation model corresponding to the first lead mode, and a first predicted signal is obtained by the ECG signal generation model; the first predicted signal is used to characterize the ECG signal predicted by the ECG signal generation model under lead modes other than the first lead mode for the sample personnel; wherein, the first lead mode is any lead mode;

[0011] The training loss value is determined based on the first predicted signal and the multi-lead ECG signal.

[0012] Based on the loss value, the parameters of the ECG signal generation model are updated to obtain the ECG signal generation model trained under the first lead mode;

[0013] The performance of the ECG signal generation model trained under various lead modes is evaluated, the performance index corresponding to the trained ECG signal generation model is determined, the optimal lead mode is determined based on the performance index, and the ECG signal generation model trained under the optimal lead mode is determined as the target model.

[0014] Acquire the second electrocardiogram signal of the target person under the optimal lead configuration;

[0015] The second electrocardiogram (ECG) signal is input into the target model, and a second prediction signal is obtained through the target model prediction; the second prediction signal is used to characterize the ECG signal of the target person in lead modes other than the optimal lead mode predicted by the target model.

[0016] The second ECG signal and the second predicted signal are spliced ​​together to obtain the generated target ECG signal.

[0017] In addition, the electrocardiogram signal generation method according to the above embodiments of this application may also have the following additional technical features:

[0018] Furthermore, in one embodiment of this application, the electrocardiogram signal generation model is constructed using a convolutional autoencoder.

[0019] Furthermore, in one embodiment of this application, the convolutional autoencoder includes an encoder and a decoder;

[0020] The encoder includes a convolutional layer, a first batch normalization layer, and a first activation function;

[0021] The decoder includes a deconvolution layer, a second batch normalization layer, and a second activation function.

[0022] Furthermore, in one embodiment of this application, after acquiring the multi-lead electrocardiogram signals corresponding to multiple sample individuals, the method further includes:

[0023] Baseline drift noise in the multi-lead ECG signal was removed using wavelet decomposition and reconstruction.

[0024] Power frequency interference and electromyographic interference in the multi-lead electrocardiogram signal were removed by wavelet thresholding.

[0025] Furthermore, in one embodiment of this application, the step of evaluating the performance of the ECG signal generation model trained under various lead modes, determining the performance index corresponding to the trained ECG signal generation model, and determining the optimal lead mode based on the performance index includes:

[0026] The index values ​​of the first predicted signal and the multi-lead ECG signal are determined by at least one of the following index algorithms: mean square error, mean absolute error, or Pearson correlation coefficient.

[0027] The lead mode with the smallest mean square error and the smallest mean absolute error is determined as the optimal lead mode, or the lead mode with the largest Pearson correlation coefficient is determined as the optimal lead mode.

[0028] Furthermore, in one embodiment of this application, the step of evaluating the performance of the ECG signal generation model trained under various lead modes, determining the performance index corresponding to the trained ECG signal generation model, and determining the optimal lead mode based on the performance index includes:

[0029] Obtain cardiovascular disease label information corresponding to each of the sample individuals; the cardiovascular disease label information is used to characterize the true category of cardiovascular disease possessed by the sample individuals;

[0030] Based on the first predicted signal output by the ECG signal generation model trained under each lead mode, the cardiovascular disease of the sample personnel is predicted by the cardiovascular disease prediction model to obtain the first prediction result.

[0031] Based on the first prediction result and the cardiovascular disease label information, the prediction accuracy corresponding to each of the lead methods is determined;

[0032] The lead with the highest prediction accuracy is determined as the optimal lead.

[0033] On the other hand, embodiments of this application provide an electrocardiogram (ECG) signal generation device, the device comprising:

[0034] The first acquisition unit is used to acquire multi-lead electrocardiogram (ECG) signals corresponding to multiple sample individuals; wherein, the multi-lead ECG signals include first ECG signals measured in various lead modes, and there are multiple types of lead modes;

[0035] The first prediction unit is used to input the first electrocardiogram (ECG) signal corresponding to the first lead mode into the ECG signal generation model corresponding to the first lead mode, and to predict the first prediction signal through the ECG signal generation model; the first prediction signal is used to characterize the ECG signal predicted by the ECG signal generation model under the lead mode other than the first lead mode of the sample person; wherein, the first lead mode is any lead mode;

[0036] The processing unit is used to determine the training loss value based on the first predicted signal and the multi-lead electrocardiogram signal.

[0037] The update unit is used to update the parameters of the ECG signal generation model according to the loss value, so as to obtain the ECG signal generation model trained in the first lead mode.

[0038] The evaluation unit is used to evaluate the performance of the ECG signal generation model trained under various lead modes, determine the performance index corresponding to the trained ECG signal generation model, determine the optimal lead mode based on the performance index, and determine the ECG signal generation model trained under the optimal lead mode as the target model.

[0039] The second acquisition unit is used to acquire the second electrocardiogram signal measured by the target person under the optimal lead mode;

[0040] The second prediction unit is used to input the second electrocardiogram signal into the target model and obtain a second prediction signal through the target model; the second prediction signal is used to characterize the electrocardiogram signal of the target person in lead modes other than the optimal lead mode predicted by the target model;

[0041] An integration unit is used to splice the second electrocardiogram signal and the second predicted signal to obtain the generated target electrocardiogram signal.

[0042] On the other hand, embodiments of this application provide a cardiovascular disease diagnostic device, including:

[0043] The input unit is used to input the target electrocardiogram signal generated by the aforementioned electrocardiogram signal generation method into the cardiovascular disease prediction model.

[0044] The third prediction unit is used to predict the cardiovascular diseases of the target individuals through a cardiovascular disease prediction model, and obtain the second prediction result.

[0045] On the other hand, embodiments of this application provide an electronic device, including:

[0046] At least one processor;

[0047] At least one memory for storing at least one program;

[0048] When the at least one program is executed by the at least one processor, the at least one processor implements the above-described electrocardiogram signal generation method.

[0049] On the other hand, embodiments of this application also provide a computer-readable storage medium storing a processor-executable program, which, when executed by a processor, is used to implement the above-described electrocardiogram signal generation method.

[0050] The advantages and beneficial effects of this application 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 this application:

[0051] The electrocardiogram (ECG) signal generation method, cardiovascular disease diagnostic device, equipment, and medium disclosed in this application acquire multi-lead ECG signals corresponding to multiple sample individuals. The multi-lead ECG signals include first ECG signals measured in various lead modes, and there are multiple types of lead modes. The first ECG signal corresponding to the first lead mode is input into an ECG signal generation model corresponding to the first lead mode, and a first predicted signal is obtained through the ECG signal generation model. The first predicted signal is used to characterize the ECG signal predicted by the ECG signal generation model for the sample individuals in lead modes other than the first lead mode. The first lead mode can be any lead mode. A training loss value is determined based on the first predicted signal and the multi-lead ECG signals. The ECG signal generation is then performed based on the loss value. The model is updated with parameters to obtain a trained ECG signal generation model under the first lead mode. The effectiveness of the trained ECG signal generation models under each lead mode is evaluated to determine the performance index corresponding to the trained ECG signal generation model. Based on the performance index, the optimal lead mode is determined, and the ECG signal generation model trained under the optimal lead mode is identified as the target model. A second ECG signal measured by the target person under the optimal lead mode is acquired. The second ECG signal is input into the target model, and a second predicted signal is obtained through the target model's prediction. The second predicted signal is used to characterize the ECG signal predicted by the target model under lead modes other than the optimal lead mode for the target person. The second ECG signal and the second predicted signal are concatenated to obtain the generated target ECG signal. This method can generate multi-lead ECG signals based on ECG signals acquired under one lead mode, eliminating the need for users to wear complex patches for ECG signal acquisition. It is suitable for smart wearable device applications, providing more comprehensive ECG signals and offering better practicality. Furthermore, this method can evaluate and select the optimal lead mode suitable as the basis for signal generation, improving the accuracy of the generated ECG signals and facilitating efficient and reliable monitoring of users' health status based on the generated ECG signals. Attached Figure Description

[0052] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the following description is provided with accompanying drawings of the relevant technical solutions in the embodiments of this application or the prior art. It should be understood that the accompanying drawings described below are only for the purpose of clearly illustrating some embodiments of the technical solutions in this application. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0053] Figure 1 This is a schematic diagram of the implementation environment for an electrocardiogram signal generation method provided in this application embodiment;

[0054] Figure 2 This is a flowchart illustrating an electrocardiogram (ECG) signal generation method provided in an embodiment of this application.

[0055] Figure 3 This is a schematic diagram of the structure of an electrocardiogram signal generation model provided in the embodiments of this application;

[0056] Figure 4 This is a schematic diagram of the structure of an encoder provided in an embodiment of this application;

[0057] Figure 5 This is a schematic diagram of the structure of a decoder provided in an embodiment of this application;

[0058] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0059] The present application will be further described below with reference to the accompanying drawings and specific embodiments. The described embodiments should not be considered as limitations on the present application, and all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of the present application.

[0060] In the following description, references are made to “some embodiments,” which describe a subset of all possible embodiments. However, it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0061] Unless otherwise defined, 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 application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0062] 1) Artificial Intelligence (AI) is the theory, methods, technology, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to achieve optimal results. In other words, AI is a comprehensive technology within computer science that attempts to understand the essence of intelligence and produce a new kind of intelligent machine that can react in a way similar to human intelligence. AI studies the design principles and implementation methods of various intelligent machines, enabling them to have perception, reasoning, and decision-making capabilities. AI technology is a comprehensive discipline involving a wide range of fields, encompassing both hardware and software technologies. Fundamental AI technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operating / interactive systems, and mechatronics. AI software technologies mainly include computer vision, speech processing, natural language processing, and machine learning / deep learning.

[0063] 2) Machine Learning (ML) is a multidisciplinary field involving probability theory, statistics, approximation theory, convex analysis, and algorithm complexity theory. It specifically studies how computers can simulate or implement human learning behavior to acquire new knowledge or skills and reorganize existing knowledge structures to continuously improve their performance. Machine learning is the core of artificial intelligence and the fundamental way to endow computers with intelligence. Its applications span all areas of artificial intelligence. Machine learning (deep learning) typically includes techniques such as artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, and instructional learning.

[0064] Cardiovascular disease is one of the leading causes of death worldwide. Electrocardiography (ECG) is commonly used for the early diagnosis of cardiovascular disease and plays a crucial role in reducing patient mortality. Healthcare professionals can diagnose cardiovascular disease by analyzing typical changes in ECGs, such as abnormalities in characteristic waves like the P wave, QRS complex, and ST segment.

[0065] Currently, electrocardiograms (ECGs) commonly used in medicine include multi-lead ECG signals, which can comprehensively reflect the health status of the heart and help doctors make accurate clinical diagnoses. However, acquiring standard multi-lead ECG signals is difficult; for example, acquiring 12-lead ECG signals requires numerous patches, restricting the patient's freedom of movement. This is only suitable for clinical medical diagnosis and is not conducive to long-term cardiovascular disease monitoring. To achieve long-term monitoring of cardiovascular diseases, everyday wearable ECG signal measurement devices (such as smart bracelets and patches) have been developed and used. These everyday wearable devices can monitor ECG signals in real time while minimizing the impact on patients' normal lives. However, everyday wearable ECG signal measurement devices have a significant limitation: most of them can only measure ECG signals from a single lead, resulting in lower practicality and difficulty in comprehensively reflecting the health status of the heart.

[0066] In view of this, this application provides a method for generating electrocardiogram (ECG) signals. This method can generate multi-lead ECG signals based on ECG signals acquired under one lead mode, eliminating the need for users to wear complex patches for ECG signal acquisition. It is suitable for application scenarios of smart wearable devices, can obtain more comprehensive ECG signals, and has better practicality. Moreover, this method can evaluate and select the optimal lead mode suitable as the basis for signal generation, which can improve the accuracy of the generated ECG signals and facilitate efficient and reliable monitoring of the user's health status based on the generated ECG signals.

[0067] Please refer to Figure 1 , Figure 1 This diagram illustrates an implementation environment for an electrocardiogram (ECG) signal generation method provided in this embodiment. In this implementation environment, the main hardware and software components involved include a terminal device 110 and a backend server 120. The terminal device 110 and the backend server 120 are connected via communication.

[0068] Specifically, the ECG signal generation method provided in this application embodiment can be executed separately on the terminal device 110, separately on the backend server 120, or based on data interaction between the terminal device 110 and the backend server 120.

[0069] The terminal device 110 in the above embodiments may include mobile phones, computers, smart wearable devices, PDA devices, smart voice interaction devices, vehicle terminals, etc., but is not limited to these. The backend server 120 may be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms.

[0070] The terminal device 110 and the backend server 120 can establish a communication connection via a wireless network or a wired network. This wireless or wired network uses standard communication technologies and / or protocols. The network can be the Internet or any other network, including but not limited to any combination of Local Area Network (LAN), Metropolitan Area Network (MAN), Wide Area Network (WAN), mobile, wired or wireless networks, private networks, or virtual private networks.

[0071] Of course, this is understandable. Figure 1 The implementation environment described in this application is only one of the optional application scenarios for the ECG signal generation method provided in this embodiment. The actual application is not fixed. Figure 1 The software and hardware environment shown.

[0072] Below, in conjunction with the foregoing description of the implementation environment, a method for generating electrocardiogram signals provided in the embodiments of this application will be introduced and explained.

[0073] Please refer to Figure 2 , Figure 2 This is a schematic diagram of an electrocardiogram (ECG) signal generation method provided in an embodiment of this application. The ECG signal generation method includes, but is not limited to:

[0074] Step 210: Obtain multi-lead electrocardiogram (ECG) signals corresponding to multiple sample individuals; wherein, the multi-lead ECG signals include the first ECG signals measured in various lead modes, and there are multiple types of lead modes;

[0075] Step 220: Input the first ECG signal corresponding to the first lead mode into the ECG signal generation model corresponding to the first lead mode, and obtain a first prediction signal through the ECG signal generation model; the first prediction signal is used to characterize the ECG signal predicted by the ECG signal generation model under lead modes other than the first lead mode for the sample person; wherein, the first lead mode is any lead mode;

[0076] Step 230: Determine the training loss value based on the first predicted signal and the multi-lead ECG signal;

[0077] Step 240: Update the parameters of the ECG signal generation model according to the loss value to obtain the ECG signal generation model trained in the first lead mode;

[0078] Step 250: Evaluate the performance of the ECG signal generation model trained under each lead mode, determine the performance index corresponding to the trained ECG signal generation model, determine the optimal lead mode based on the performance index, and determine the ECG signal generation model trained under the optimal lead mode as the target model.

[0079] Step 260: Obtain the second electrocardiogram signal of the target person under the optimal lead mode;

[0080] Step 270: Input the second electrocardiogram signal into the target model, and obtain a second prediction signal through the target model; the second prediction signal is used to characterize the electrocardiogram signal of the target person in lead modes other than the optimal lead mode predicted by the target model;

[0081] Step 280: The second ECG signal and the second predicted signal are spliced ​​together to obtain the generated target ECG signal.

[0082] This application provides a method for generating electrocardiogram (ECG) signals. This method can automatically generate multi-lead ECG signals based on an ECG signal generation model. In actual use, it is only necessary to collect the ECG signal of the user in one lead mode, without requiring the user to wear a complicated patch for ECG signal collection. This method is suitable for application scenarios of smart wearable devices.

[0083] In this embodiment of the application, before generating the electrocardiogram (ECG) signal, it is necessary to train and determine the ECG signal generation model to be used. In this embodiment of the application, the ECG signal generation model that is finally put into use is denoted as the target model. The process of determining the target model will be introduced and explained below.

[0084] Specifically, when determining the target model, multi-lead ECG signals corresponding to multiple sample individuals can be acquired first. Here, sample individuals are those whose relevant data are used when training and evaluating the ECG signal generation model. These individuals can be patients who have undergone ECG testing in hospitals, etc. The number of sample individuals can be multiple, and this application does not limit the specific number. In the embodiments of this application, to facilitate subsequent processing, the number of leads and the specific lead method for acquiring the multi-lead ECG signals corresponding to each sample individual can be the same. For example, in some embodiments, a standard 12-lead ECG signal can be acquired as the multi-lead ECG signal. Here, the standard 12-lead ECG signal captures the heart's electrical signals by placing electrodes at different points on the body. There are 12 lead settings, such as: Standard Lead I: left forearm to right forearm; Standard Lead II: right forearm to left lower limb; Standard Lead III: left lower limb to right forearm; Pressure Limb Lead aVR: right upper limb (positive) to left upper limb and left lower limb (negative); Pressure Limb Lead aVL: left upper limb (positive) to right upper limb and left lower limb. (Negative pole); Pressure limb lead aVF: Left lower limb (positive pole) to right upper limb and left upper limb (negative pole); Chest lead V1: at the right sternal border in the fourth intercostal space; Chest lead V2: at the left sternal border in the fourth intercostal space; Chest lead V3: between V2 and V4 (approximately at the midline of the sternum); Chest lead V4: at the intersection of the left midclavicular line and the fifth intercostal space; Chest lead V5: at the same level as V4 at the left anterior axillary line; Chest lead V6: at the same level as V4 at the left midaxillary line.

[0085] In this embodiment of the application, the ECG signals measured in each lead mode of the multi-lead ECG signal are recorded as the first ECG signal. It is understood that in this embodiment of the application, the multi-lead ECG signal may also include only the first ECG signal in a portion of the lead modes of the standard 12-lead ECG signal. This application does not limit the specific number and type of lead modes.

[0086] Of course, it should be noted that when obtaining multi-lead electrocardiogram signals from sample personnel, relevant laws and regulations must be followed, and the privacy of the sample personnel must not be disclosed.

[0087] In this embodiment, after obtaining the first electrocardiogram (ECG) signals measured under various lead methods of the sample personnel, the ECG signals can be divided according to the lead method, and multiple ECG signal generation models can be trained. Here, the ECG signal generation model can be used to generate ECG signals, and its input can be the ECG signal measured under a certain lead method, while the output is the predicted ECG signals corresponding to other lead methods.

[0088] Specifically, in this embodiment, for each lead mode, during the training of its corresponding ECG signal generation model, it can be designated as the first lead mode. The first ECG signal corresponding to the first lead mode is input into the ECG signal generation model corresponding to the first lead mode. The ECG signal generation model can predict a predicted signal, which in this embodiment is designated as the first predicted signal. The first predicted signal is the ECG signal of the sample personnel predicted by the ECG signal generation model under lead modes other than the first lead mode. For example, assuming the current multi-lead ECG signal is a 12-lead signal, one lead mode is first determined as the first lead mode. The first ECG signal under the first lead mode can be used as input, and the ECG signal generation model can predict the ECG signals under the other 11 lead modes. Then, based on the predicted first predicted signal and the original multi-lead ECG signal, the prediction effect of the ECG signal generation model can be determined, that is, the loss value between the first predicted signal and the multi-lead ECG signal can be calculated using a relevant loss function. This application does not limit the specific type of loss function used. Based on the loss value, the parameters of the ECG signal generation model can be updated to obtain the trained ECG signal generation model in the first lead mode.

[0089] For example, in some scenarios, the Adam optimizer can be used to train the model with Mean Squared Error (MSE) as the loss function. The initial learning rate is set to 0.01, and an early stopping strategy is adopted. If the loss function value on the validation set does not improve for three consecutive epochs, the learning rate is reduced to one-tenth of the previous value; if the loss function on the validation set does not improve for eight consecutive epochs, model training is stopped. The batch size of the model is 128, and the maximum number of iterations is 100.

[0090] It is understood that in the embodiments of this application, each lead mode is sequentially used as the first lead mode, and an ECG signal generation model corresponding to each lead mode can be trained. After obtaining the trained ECG signal generation models under each lead mode, their effectiveness can be evaluated to determine the performance index corresponding to each model. Based on the performance index, the optimal lead mode can be selected from each lead mode, and the ECG signal generation model trained under the optimal lead mode can be determined as the target model. For example, in some embodiments, the ECG signal generation model with the highest accuracy can be evaluated and determined as the target model. This will be described in detail in subsequent embodiments and will not be elaborated here.

[0091] In this embodiment, after determining the optimal lead configuration and target model, the application of ECG signal generation can be implemented based on the optimal lead configuration and target model. Specifically, in this embodiment, the personnel who actually need to generate ECG signals can be designated as the target personnel. The ECG signal measured by the target personnel under the optimal lead configuration can be obtained and designated as the second ECG signal. Then, the second ECG signal can be input into the target model, and the target model can predict the corresponding ECG signals of the target personnel under lead configurations other than the optimal lead configuration. These signals are designated as the second predicted signal. By concatenating the second ECG signal and the second predicted signal, target ECG signals under multiple lead configurations can be obtained.

[0092] It is understood that the ECG signal generation method provided in this application embodiment can generate multi-lead ECG signals based on ECG signals acquired under one lead mode, without requiring users to wear complex patches for ECG signal acquisition. It is suitable for application scenarios of smart wearable devices, can obtain more comprehensive ECG signals, and has better practicality. Moreover, this method can evaluate and select the optimal lead mode suitable as the basis for signal generation, which can improve the accuracy of the generated ECG signals and facilitate efficient and reliable monitoring of the user's health status based on the generated ECG signals.

[0093] Specifically, in some embodiments, the electrocardiogram signal generation model is built using a convolutional autoencoder.

[0094] In this embodiment, a convolutional autoencoder can be used to build an electrocardiogram (ECG) signal generation model. Specifically, please refer to... Figure 3 , Figure 3 This illustration shows a schematic diagram of a convolutional autoencoder (CA) provided in an embodiment of this application. In this embodiment, the CA is a neural network structure that includes an encoder and a decoder. The encoder compresses the input signal into a feature representation in the latent space, and the decoder reconstructs the input signal from the feature representation in the latent space. Alternatively, the decoder can generate a signal of the same type as the input signal based on the feature representation in the latent space. For the application in this embodiment, an ECG signal measured in a single-lead configuration can be used as input, and the CA can generate ECG signals in multiple other lead configurations using the CA. Figure 1 As shown in the embodiment of this application, the convolutional autoencoder can first generate the remaining 11-lead ECG signals based on the input single-lead ECG signal, and then splice the input real single-lead ECG signal and the generated 11-lead ECG signal to form a standard 12-lead ECG signal.

[0095] Reference Figure 4 and Figure 5 ,in Figure 4This is a schematic diagram of the structure of an encoder provided in an embodiment of this application. Figure 5 This is a schematic diagram of a decoder structure provided in an embodiment of this application. In this embodiment, the encoder within the convolutional autoencoder may include a convolutional layer (Conv), a batch normalization layer (BN), and an activation function, wherein the batch normalization layer and the activation function are denoted as the first batch normalization layer and the first activation function. Specifically, the convolutional layer (Conv) can be used to learn the feature information of the electrocardiogram (ECG) signal. Compared with fully connected layers, the convolutional layer has fewer parameters and is more likely to extract local invariant features from the ECG signal. In this embodiment, the convolutional layer is used to learn the feature information in the original single-lead ECG signal and compress it into a feature representation in the latent space. Since the ECG signal data is one-dimensional data, the convolutional layer is also designed as a one-dimensional convolution. The number of output channels of the five convolutional layers are 16, 32, 64, 128, and 256, respectively, and the kernel size is set to 3, with a stride of 2.

[0096] The first batch normalization layer (BN) normalizes each batch of data, making its mean close to 0 and its variance close to 1. This helps address the internal covariate shift problem, making network training more stable and improving the network's generalization ability. The first activation function can be the LeakyReLU function, a non-linear function that enhances the model's expressive power and allows it to better fit non-linear data. Furthermore, compared to the ReLU activation function, the LeakyReLU function sets a non-zero slope for the negative part, avoiding neuron death. In this embodiment, the slope of the LeakyReLU function for the negative part is set to 0.2.

[0097] In this embodiment, the decoder includes a deconvolutional layer (ConvTranspose), a batch normalization layer (BN), and an activation function. The batch normalization layer and activation function are referred to as the second batch normalization layer and the second activation function, respectively. Specifically, the deconvolutional layer (ConvTranspose) is an upsampling method. Unlike the convolutional layer, it increases the dimensionality of feature information, used to upsample and generate new ECG signals from the feature information in the latent space. In this embodiment, the deconvolutional layer can generate the ECG signals of the remaining 11 leads based on the feature representation of the input single-lead ECG signal in the latent space. Since the ECG signal data is one-dimensional, the deconvolutional layer is also designed as a one-dimensional deconvolution. The number of output channels for the five deconvolutional layers are 128, 64, 32, 16, and 11, respectively. The kernel size is set to 3, and the stride is 2. The second activation function can still be the LeakyReLU function. The parameter settings and functions of the second batch normalization layer (BN) and the second activation function are the same as those of the encoder, and will not be repeated here.

[0098] Specifically, in some embodiments, after acquiring the multi-lead electrocardiogram signals corresponding to multiple sample individuals, the method further includes:

[0099] Baseline drift noise in the multi-lead ECG signal was removed using wavelet decomposition and reconstruction.

[0100] Power frequency interference and electromyographic interference in the multi-lead electrocardiogram signal were removed by wavelet thresholding.

[0101] It should be noted that the original electrocardiogram (ECG) signals are subject to various noise interferences during the acquisition process. These noises can adversely affect the subsequent training of the model and the evaluation of the model's performance. Therefore, in this embodiment, after the ECG signals are acquired, they can be denoised.

[0102] Specifically, common noise interference in raw electrocardiogram (ECG) signals includes baseline drift noise, power line interference, and electromyographic interference. Baseline drift noise is generally generated by the subject's breathing or movement and is a low-frequency signal, with a frequency range of approximately 0.05–1 Hz. Baseline drift noise can be removed using wavelet decomposition and reconstruction. Taking a sampling frequency of 100 Hz as an example, the ECG signal can be decomposed into 6 levels using the db6 wavelet basis function. Baseline drift noise is mainly located in the frequency band corresponding to the 6th level after wavelet decomposition, so the approximation coefficients and detail coefficients corresponding to the 6th level can be forcibly set to zero, thereby filtering out baseline drift noise in the ECG signal.

[0103] Power frequency interference, caused by power lines and surrounding electrical equipment, is a high-frequency noise in the range of 50–60 Hz. Electromyographic interference, caused by muscle contraction and relaxation, is a high-frequency noise in the range of 5–500 Hz. Both types of noise can be removed using wavelet thresholding. The ECG signal is decomposed into first-order components using the db6 wavelet basis function. An unbiased risk estimation method is used to calculate the threshold, and this threshold is then used to perform soft-threshold denoising on the detail coefficients at each scale obtained after decomposition, resulting in the denoised ECG signal.

[0104] Specifically, in some embodiments, the step of evaluating the performance of the ECG signal generation model trained under various lead modes, determining the performance index corresponding to the trained ECG signal generation model, and determining the optimal lead mode based on the performance index includes:

[0105] The index values ​​of the first predicted signal and the multi-lead ECG signal are determined by at least one of the following index algorithms: mean square error, mean absolute error, or Pearson correlation coefficient.

[0106] The lead mode with the smallest mean square error and the smallest mean absolute error is determined as the optimal lead mode, or the lead mode with the largest Pearson correlation coefficient is determined as the optimal lead mode.

[0107] In this embodiment, when evaluating the performance of the ECG signal generation model, in some scenarios, metrics such as mean squared error (MSE), mean absolute error (MAE), and Pearson correlation coefficient can be used for evaluation. The smaller the MSE and MAE, and the larger the Pearson correlation coefficient, the better the model's performance. Specifically, please refer to Table 1 below. In this embodiment, taking the task of generating a complete 12-lead ECG signal from a single-lead ECG signal as an example, ECG signal generation models corresponding to different lead configurations were trained, and their performance metrics were evaluated. Table 1 shows some of the obtained metric values.

[0108] Table 1

[0109]

[0110]

[0111]

[0112] As shown in Table 1, when using ECG signals from 12 different leads as input, the average mean squared error (MSE) of the model-generated ECG signals in the remaining 11 leads compared to the actual signals is below 0.02, the average mean absolute error (MAO) is below 0.06, and the average Pearson correlation coefficient is above 0.7. Specifically, when lead V3 is used as input, the average MSE and MAO of the model-generated ECG signals in the remaining 11 leads compared to the actual signals are the smallest, at 0.011 and 0.050, respectively; when lead III is used as input, the average MSE and MAO of the model-generated ECG signals in the remaining 11 leads compared to the actual signals are the largest, at 0.018 and 0.059, respectively. When lead II is used as input, the average Pearson correlation coefficient of the remaining 11 leads generated by the model is the largest compared to the actual signal, at 0.799. When lead V1 is used as input, the average Pearson correlation coefficient of the remaining 11 leads generated by the model is the smallest compared to the actual signal, at 0.712. In some embodiments, the lead mode with the smallest mean square error and mean absolute error can be determined as the optimal lead mode. For example, lead V3 can be determined as the optimal lead mode, and the ECG signal generation model trained under lead V3 can be determined as the target model. Of course, in some embodiments, the lead mode with the largest Pearson correlation coefficient can also be determined as the optimal lead mode. For example, lead II can be determined as the optimal lead mode, and the ECG signal generation model trained under lead II can be determined as the target model. In other embodiments, different index values ​​can be weighted to determine the optimal lead mode, and this application does not limit this.

[0113] Specifically, in some embodiments, the step of evaluating the performance of the ECG signal generation model trained under various lead modes, determining the performance index corresponding to the trained ECG signal generation model, and determining the optimal lead mode based on the performance index includes:

[0114] Obtain cardiovascular disease label information corresponding to each of the sample individuals; the cardiovascular disease label information is used to characterize the true category of cardiovascular disease possessed by the sample individuals;

[0115] Based on the first predicted signal output by the ECG signal generation model trained under each lead mode, the cardiovascular disease of the sample personnel is predicted by the cardiovascular disease prediction model to obtain the first prediction result.

[0116] Based on the first prediction result and the cardiovascular disease label information, the prediction accuracy corresponding to each of the lead methods is determined;

[0117] The lead with the highest prediction accuracy is determined as the optimal lead.

[0118] In this embodiment, the generated electrocardiogram (ECG) signal can generally be used in a cardiovascular disease prediction model. To improve the practicality of the target ECG signal generated in this embodiment, the ECG signal generation model can also be evaluated by combining it with a subsequent classification and prediction task.

[0119] Specifically, in this embodiment, a cardiovascular disease prediction model can be used to predict the cardiovascular diseases present in a person. The input of the cardiovascular disease prediction model is an electrocardiogram (ECG) signal, and the output is the predicted category of the cardiovascular disease. In this embodiment, cardiovascular disease label information corresponding to the sample person can be obtained, which can be used to characterize the true category of the cardiovascular disease present in the sample person. For example, the true category of cardiovascular disease may include healthy, myocardial infarction, ST-T segment changes, myocardial hypertrophy, conduction disorder, etc., and this application does not limit this. After training each ECG signal generation model, the cardiovascular disease present in the sample person can be predicted by the cardiovascular disease prediction model based on the first predicted signal output by the trained ECG signal generation model and the input of the ECG signal generation model (the first ECG signal corresponding to the first lead mode), resulting in a prediction result, denoted as the first prediction result. The first prediction result can characterize the category of cardiovascular disease present in the sample person predicted by the cardiovascular disease prediction model. In this embodiment, the prediction accuracy corresponding to each lead mode can be determined based on the first prediction result and the cardiovascular disease label information.

[0120] To verify the practicality of the ECG signal generation model in cardiovascular disease diagnosis, a disease diagnosis classification experiment was conducted on the PTB database using a residual network as the classifier. The PTB database is a diagnostic database related to myocardial infarction, a common cardiovascular disease. In this embodiment, the classification performance evaluation metrics selected during classification can be accuracy (Acc) and F1 score (F1), defined as follows:

[0121]

[0122] TP, TN, FP, and FN represent the number of true positives, true negatives, false positives, and false negatives, respectively.

[0123] For ease of evaluation, this application sets up two scenarios and compares them. Scenario (1): Disease diagnosis and classification are performed directly using single-lead ECG signals; Scenario (2): A pre-trained ECG signal generation model is first used to generate a complete 12-lead ECG signal based on the single-lead ECG signal, and then the complete ECG signal is used for diagnosis and classification. During the training process, the parameters of the convolutional autoencoder are frozen, and only the parameters of the residual network classifier are trained. The results are shown in Table 2.

[0124] Table 2

[0125]

[0126] As shown in Table 2, except for Lead aVF and Lead V1, the performance of the other 10 leads in disease diagnosis and classification was improved after the pre-trained ECG signal generation model generated complete 12-lead signals, compared to directly using single-lead ECG signals. When Lead I was used as the input signal, the application of the ECG signal generation model significantly improved the performance of disease diagnosis and classification, with accuracy and F1 score improvements of 9.4% and 10.64%, respectively. When Lead V6 was used as the input signal, the model performed best in disease diagnosis and classification, with accuracy and F1 score of 43.04% and 37.37% without the convolutional autoencoder (case 1), and 48.21% and 42.88% with the convolutional autoencoder (case 2). This indicates that the application of the ECG signal generation model can improve the overall accuracy of the model in diagnosing myocardial infarction and has certain practical value. Therefore, lead V6 can be determined as the optimal lead.

[0127] Based on comprehensive analysis, when diagnosing myocardial infarction, a typical cardiovascular disease, the V6 lead can be selected as the acquisition lead. By using an ECG signal generation model, a complete 12-lead signal can be generated from the V6 lead signal, which can further improve the accuracy of the model in diagnosing myocardial infarction.

[0128] In this embodiment of the application, after obtaining the target person's corresponding target electrocardiogram signal, it can also be input into the cardiovascular disease prediction model. The cardiovascular disease prediction model can then be used to predict the cardiovascular disease of the target person and obtain the prediction result, which is recorded as the second prediction result.

[0129] In this embodiment of the application, an electrocardiogram (ECG) signal generation device is also provided, comprising:

[0130] The first acquisition unit is used to acquire multi-lead electrocardiogram (ECG) signals corresponding to multiple sample individuals; wherein, the multi-lead ECG signals include first ECG signals measured in various lead modes, and there are multiple types of lead modes;

[0131] The first prediction unit is used to input the first electrocardiogram (ECG) signal corresponding to the first lead mode into the ECG signal generation model corresponding to the first lead mode, and to predict the first prediction signal through the ECG signal generation model; the first prediction signal is used to characterize the ECG signal predicted by the ECG signal generation model under lead modes other than the first lead mode for the sample personnel; wherein, the first lead mode is any lead mode;

[0132] The processing unit is used to determine the training loss value based on the first predicted signal and the multi-lead electrocardiogram signal.

[0133] The update unit is used to update the parameters of the ECG signal generation model according to the loss value, so as to obtain the ECG signal generation model trained in the first lead mode.

[0134] The evaluation unit is used to evaluate the performance of the ECG signal generation model trained under various lead modes, determine the performance index corresponding to the trained ECG signal generation model, determine the optimal lead mode based on the performance index, and determine the ECG signal generation model trained under the optimal lead mode as the target model.

[0135] The second acquisition unit is used to acquire the second electrocardiogram signal measured by the target person under the optimal lead mode;

[0136] The second prediction unit is used to input the second electrocardiogram signal into the target model and obtain a second prediction signal through the target model; the second prediction signal is used to characterize the electrocardiogram signal of the target person in lead modes other than the optimal lead mode predicted by the target model;

[0137] An integration unit is used to splice the second electrocardiogram signal and the second predicted signal to obtain the generated target electrocardiogram signal.

[0138] This application embodiment also provides a cardiovascular disease diagnostic device, comprising:

[0139] The input unit is used to input the target electrocardiogram signal generated by the aforementioned electrocardiogram signal generation method into the cardiovascular disease prediction model.

[0140] The third prediction unit is used to predict the cardiovascular diseases of the target individuals through a cardiovascular disease prediction model, and obtain the second prediction result.

[0141] It is understood that the content of the above method embodiments is applicable to this system embodiment. The specific functions implemented in this system embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0142] Reference Figure 6 This application provides an electronic device, including:

[0143] At least one processor 610;

[0144] At least one memory 620 is used to store at least one program;

[0145] When at least one program is executed by at least one processor 610, the at least one processor 610 implements the above-described electrocardiogram signal generation method.

[0146] Similarly, the content of the above method embodiments is applicable to the embodiments of this electronic device. The specific functions implemented by the embodiments of this electronic device are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0147] This application embodiment also provides a computer-readable storage medium storing a program executable by a processor 610, which, when executed by the processor 610, is used to perform the above-described electrocardiogram signal generation method.

[0148] Similarly, the content of the above method embodiments is applicable to the present computer-readable storage medium embodiments. The specific functions implemented by the present computer-readable storage medium embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0149] In some alternative embodiments, the functions / operations mentioned in the block diagrams may not occur in the order shown in the operation diagrams. For example, depending on the functions / operations involved, two consecutively shown blocks may actually be executed substantially simultaneously, or the blocks may sometimes be executed in reverse order. Furthermore, the embodiments presented and described in the flowcharts of this application are provided by way of example to provide a more comprehensive understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is changed and sub-operations described as part of a larger operation are executed independently.

[0150] Furthermore, although this application is described in the context of functional modules, it should be understood that, unless otherwise stated to the contrary, one or more of the functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in a separate physical device or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding this application. Rather, given the properties, functions, and internal relationships of the various functional modules in the apparatus disclosed herein, the actual implementation of the module will be understood within the scope of conventional technology for an engineer. Therefore, those skilled in the art can implement the application set forth in the claims using ordinary techniques without excessive experimentation. It is also understood that the specific concepts disclosed are merely illustrative and not intended to limit the scope of this application, which is determined by the full scope of the appended claims and their equivalents.

[0151] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0152] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0153] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which programs can be printed, because programs can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0154] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0155] In the foregoing description of this specification, the references to terms such as "one embodiment," "another embodiment," or "some embodiments," etc., indicate that a specific feature, structure, material, or characteristic described in connection with an embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0156] Although embodiments of this application have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of this application, the scope of which is defined by the claims and their equivalents.

[0157] The above is a detailed description of the preferred embodiments of this application, but this application is not limited to the embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of this application, and these equivalent modifications or substitutions are all included within the scope defined by the claims of this application.

Claims

1. A method for generating electrocardiogram (ECG) signals, characterized in that, The method includes: Acquire multi-lead electrocardiogram (ECG) signals corresponding to multiple sample individuals; wherein, the multi-lead ECG signals include the first ECG signals measured in various lead modes, and there are multiple types of lead modes; The first electrocardiogram (ECG) signal corresponding to the first lead mode is input into the ECG signal generation model corresponding to the first lead mode, and a first predicted signal is obtained by the ECG signal generation model; the first predicted signal is used to characterize the ECG signal predicted by the ECG signal generation model under lead modes other than the first lead mode for the sample personnel; wherein, the first lead mode is any lead mode; The training loss value is determined based on the first predicted signal and the multi-lead ECG signal. Based on the loss value, the parameters of the ECG signal generation model are updated to obtain the ECG signal generation model trained under the first lead mode; The performance of the ECG signal generation model trained under various lead modes is evaluated, the performance index corresponding to the trained ECG signal generation model is determined, the optimal lead mode is determined based on the performance index, and the ECG signal generation model trained under the optimal lead mode is determined as the target model. Acquire the second electrocardiogram signal of the target person under the optimal lead configuration; The second electrocardiogram (ECG) signal is input into the target model, and a second prediction signal is obtained through the target model prediction; the second prediction signal is used to characterize the ECG signal of the target person in lead modes other than the optimal lead mode predicted by the target model. The second ECG signal and the second predicted signal are spliced ​​together to obtain the generated target ECG signal; The step of evaluating the performance of the trained ECG signal generation model under various lead modes, determining the performance index corresponding to the trained ECG signal generation model, and determining the optimal lead mode based on the performance index includes: Obtain cardiovascular disease label information corresponding to each of the sample individuals; the cardiovascular disease label information is used to characterize the true category of cardiovascular disease possessed by the sample individuals; Based on the first predicted signal output by the ECG signal generation model trained under each lead mode, the cardiovascular disease of the sample personnel is predicted by the cardiovascular disease prediction model to obtain the first prediction result. Based on the first prediction result and the cardiovascular disease label information, the prediction accuracy corresponding to each of the lead methods is determined; The lead with the highest prediction accuracy is determined as the optimal lead.

2. The method for generating electrocardiogram signals according to claim 1, characterized in that, The electrocardiogram signal generation model is built using a convolutional autoencoder.

3. The method for generating electrocardiogram signals according to claim 2, characterized in that, The convolutional autoencoder includes an encoder and a decoder; The encoder includes a convolutional layer, a first batch normalization layer, and a first activation function; The decoder includes a deconvolution layer, a second batch normalization layer, and a second activation function.

4. The method for generating electrocardiogram signals according to claim 1, characterized in that, After acquiring the multi-lead electrocardiogram signals corresponding to multiple sample individuals, the method further includes: Baseline drift noise in the multi-lead ECG signal was removed using wavelet decomposition and reconstruction. Power frequency interference and electromyographic interference in the multi-lead electrocardiogram signal were removed by wavelet thresholding.

5. The method for generating electrocardiogram signals according to claim 1, characterized in that, The step of evaluating the performance of the trained ECG signal generation model under various lead modes, determining the performance index corresponding to the trained ECG signal generation model, and determining the optimal lead mode based on the performance index further includes: The index values ​​of the first predicted signal and the multi-lead ECG signal are determined by at least one of the following index algorithms: mean square error, mean absolute error, or Pearson correlation coefficient. The lead mode with the smallest mean square error and the smallest mean absolute error is determined as the optimal lead mode, or the lead mode with the largest Pearson correlation coefficient is determined as the optimal lead mode.

6. An electrocardiogram (ECG) signal generation device, characterized in that, The device includes: The first acquisition unit is used to acquire multi-lead electrocardiogram (ECG) signals corresponding to multiple sample individuals; wherein, the multi-lead ECG signals include first ECG signals measured in various lead modes, and there are multiple types of lead modes; The first prediction unit is used to input the first electrocardiogram (ECG) signal corresponding to the first lead mode into the ECG signal generation model corresponding to the first lead mode, and to predict the first prediction signal through the ECG signal generation model; the first prediction signal is used to characterize the ECG signal predicted by the ECG signal generation model under lead modes other than the first lead mode for the sample personnel; wherein, the first lead mode is any lead mode; The processing unit is used to determine the training loss value based on the first predicted signal and the multi-lead electrocardiogram signal. The update unit is used to update the parameters of the ECG signal generation model according to the loss value, so as to obtain the ECG signal generation model trained in the first lead mode. The evaluation unit is used to evaluate the performance of the ECG signal generation model trained under various lead modes, determine the performance index corresponding to the trained ECG signal generation model, determine the optimal lead mode based on the performance index, and determine the ECG signal generation model trained under the optimal lead mode as the target model. The second acquisition unit is used to acquire the second electrocardiogram signal measured by the target person under the optimal lead mode; The second prediction unit is used to input the second electrocardiogram signal into the target model and obtain a second prediction signal through the target model; the second prediction signal is used to characterize the electrocardiogram signal of the target person in lead modes other than the optimal lead mode predicted by the target model; An integration unit is used to splice the second ECG signal and the second predicted signal to obtain the generated target ECG signal; The evaluation unit is specifically used for: Obtain cardiovascular disease label information corresponding to each of the sample individuals; the cardiovascular disease label information is used to characterize the true category of cardiovascular disease possessed by the sample individuals; Based on the first predicted signal output by the ECG signal generation model trained under each lead mode, the cardiovascular disease of the sample personnel is predicted by the cardiovascular disease prediction model to obtain the first prediction result. Based on the first prediction result and the cardiovascular disease label information, the prediction accuracy corresponding to each of the lead methods is determined; The lead with the highest prediction accuracy is determined as the optimal lead.

7. A cardiovascular disease diagnostic device, characterized in that, The device includes: An input unit is configured to input a target electrocardiogram (ECG) signal generated by the ECG signal generation method as described in any one of claims 1-5 into a cardiovascular disease prediction model. The third prediction unit is used to predict the cardiovascular diseases of the target individuals through a cardiovascular disease prediction model, and obtain the second prediction result.

8. An electronic device, characterized in that, include: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements an electrocardiogram signal generation method as described in any one of claims 1-5.

9. A computer-readable storage medium storing a processor-executable program, characterized in that, The processor-executable program, when executed by the processor, is used to implement an electrocardiogram signal generation method as described in any one of claims 1-5.

Images