A method, system and apparatus for predicting risk of sudden cardiac death within ten years
By using the China-PAR risk assessment model and convolutional neural network to process electrocardiogram data, a risk prediction model for sudden cardiac death was constructed. This solved the problem of inaccurate prediction results in existing technologies, achieved high-precision prediction of sudden cardiac death risk, and reduced the recurrence rate and mortality rate of sudden cardiac death.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING JIAN GUAN GEM FLOWER CLINIC CO LTD
- Filing Date
- 2023-08-23
- Publication Date
- 2026-07-03
AI Technical Summary
Existing methods for predicting the risk of sudden cardiac death are based on linear assumptions, which leads to inaccurate predictions and makes it impossible to accurately determine an individual's risk of sudden death.
Using a risk assessment model based on China-PAR and a convolutional neural network, 48 hours of electrocardiogram data were acquired, processed through multi-dimensional splitting and combination, and a convolutional neural network model was constructed. The model was then trained and validated to predict the risk of sudden cardiac death, and the predicted risk level of sudden cardiac death within ten years was output.
It achieves nonlinear, high-precision prediction of sudden cardiac death risk, improves prediction accuracy, reduces the recurrence rate and mortality of sudden cardiac death, and provides a low-cost, non-invasive prediction solution.
Smart Images
Figure CN117158985B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of disease risk prediction technology, specifically to a method, system, and device for predicting the risk of sudden cardiac death over ten years. Background Technology
[0002] Sudden cardiac death is a leading cause of death. The incidence of sudden cardiac death events has increased significantly in recent years in certain occupations and positions, and the affected population is trending younger. In many cases, sudden cardiac death (SCD) may be the first manifestation of an underlying heart disease, as nearly half of SCD patients had no prior diagnosis of heart disease. Why sudden cardiac death occurs or how to determine if an individual is at risk are the research directions of this invention. Some individuals may have a low risk of sudden cardiac death; some high-risk patients do not receive timely and necessary treatment, resulting in death during their prime years.
[0003] Current methods for predicting sudden cardiac death risk utilize traditional statistical tools, assuming a linear relationship between risk factors and cardiovascular prevalence. While this has some guiding significance, the variables in these hypothetical models interact linearly and additively. In reality, the interaction between risk factors and "disease detection specificity" is non-linear, and the presence or absence of a variable determines whether other variables are included in the prediction results, directly affecting whether other variables gain or lose significance, leading to inaccurate predictions. Summary of the Invention
[0004] In view of this, the present invention provides a method, system, and device for predicting the risk of sudden cardiac death over ten years, in order to solve the problem of low accuracy in the prediction results of sudden cardiac death risk in the prior art.
[0005] In a first aspect, the present invention provides a method for predicting the risk of sudden cardiac death over ten years, comprising:
[0006] Acquire 48 hours of ECG data from a preset number of individual samples and the data required for the China-PAR risk assessment model. Generate risk grading assessment results based on the China-PAR scale based on the data collected by the China-PAR risk assessment model, and divide them into training set and test set according to a preset ratio.
[0007] The electrocardiogram data is split and combined in different dimensions to obtain a preset multi-dimensional waveform.
[0008] A convolutional neural network model is constructed, taking the pre-defined multi-dimensional waveforms of the training set as input and the corresponding risk grading assessment results based on the China-PAR scale as output. The convolutional neural network model is trained and validated using a test set to obtain a sudden cardiac death risk prediction model that meets the pre-defined requirements.
[0009] The subject's 48-hour electrocardiogram data was acquired and processed by splitting and combining data in different dimensions to obtain a preset multi-dimensional waveform. This waveform was then input into the sudden cardiac death risk prediction model, and the output results served as the prediction level of the ten-year sudden cardiac death risk.
[0010] The ten-year risk prediction method for sudden cardiac death provided in this embodiment enables the general population and physicians to create a non-linear, high-precision, iterative deep learning model for the risk of sudden cardiac death. This model can predict the risk of sudden cardiac death in a low-cost, non-invasive, and immediate manner, thereby improving the detection rate of sudden cardiac death and reducing the recurrence and mortality rates.
[0011] In one optional implementation, the electrocardiogram (ECG) data is three-lead ECG data.
[0012] The embodiments of the present invention use three-lead electrocardiogram data, which can be collected using portable, high-precision electrocardiogram acquisition and recording equipment, and can also obtain the subject's electrocardiogram data relatively comprehensively without burdening the amount of data to be processed later.
[0013] In one optional implementation, the electrocardiogram data is split and combined in different dimensions to obtain a preset multi-dimensional waveform, including:
[0014] The 48-hour ECG data from the three-lead ECG was divided into two 24-hour periods, each 24-hour period was further divided into 24 one-hour periods, and each one-hour period was further divided into 6 ten-minute periods.
[0015] Generate a three-lead waveform for the first minute of each of the ten one-minute intervals. Then, connect the three-lead waveforms for the first minute of each of the six one-minute intervals in one hour to generate a composite image. Finally, stack the composite images generated for each hour in the 24-hour period together to form a three-dimensional tensor.
[0016] Repeat the above steps to generate a three-dimensional tensor for the second minute of each of the six 10-minute intervals, until all 10 three-dimensional tensors for the 24-hour period are generated. This will result in 20 three-dimensional tensors for each individual's 48-hour ECG data.
[0017] In this embodiment of the invention, three-lead electrocardiogram (ECG) data is divided into different subdivisions and combined by superposition. This allows the neural network to learn the impact of ECG data at different time points on risk levels from different dimensions.
[0018] In one alternative implementation, the data collected based on the China-PAR risk assessment model includes: each subject's gender, age, current residence, region, waist circumference, total cholesterol, high-density lipoprotein cholesterol, current blood pressure level, whether they are taking antihypertensive medication, whether they have diabetes, whether they currently smoke, and whether they have a family history of cardiovascular disease.
[0019] By comprehensively acquiring the physiological and examination history data of the subjects, the embodiments of the present invention can more accurately obtain the risk level assessment results of the China-PAR risk assessment model.
[0020] In one alternative implementation, prior to obtaining the subject's 48-hour electrocardiogram data, the method further includes obtaining the subject's informed consent form.
[0021] This invention protects the rights of users by obtaining the authorization of the subjects through methods such as acquiring data, obtaining informed consent forms, and authorizing authorization forms.
[0022] Secondly, the present invention provides a ten-year prediction system for the risk of sudden cardiac death. The system includes: a sample data acquisition module, used to acquire 48-hour electrocardiogram data of a preset number of individual samples and data required for the China-PAR risk assessment model, and to generate a risk grading assessment result based on the China-PAR scale based on the data collected by the China-PAR risk assessment model, which is divided into a training set and a test set according to a preset ratio.
[0023] The electrocardiogram (ECG) data processing module is used to split and combine the ECG data in different dimensions to obtain a preset multi-dimensional waveform.
[0024] The module for obtaining a risk prediction model for sudden cardiac death is used to construct a convolutional neural network model. It takes a preset multi-dimensional waveform of the training set as input, the corresponding risk grading assessment result based on the China-PAR scale as output, trains the convolutional neural network model, and verifies it with a test set to obtain a risk prediction model for sudden cardiac death that meets the preset requirements.
[0025] The ten-year cardiac sudden death risk prediction module is used to acquire 48 hours of electrocardiogram data from the subject, and to perform different dimensions of splitting and combination processing to obtain a preset multi-dimensional waveform. This waveform is then input into the cardiac sudden death risk prediction model, and the output result serves as the prediction level of the ten-year cardiac sudden death risk.
[0026] The ten-year cardiac sudden death risk prediction system provided in this embodiment can enable the general population and doctors to create a non-linear, high-precision, iterative deep learning model of cardiac sudden death risk, which can predict the risk of cardiac sudden death in a low-cost, non-invasive and immediate manner, thereby improving the detection rate of cardiac sudden death and reducing the recurrence rate and mortality rate.
[0027] In one optional implementation, the system further includes a user authorization module for obtaining informed consent from the subject.
[0028] This invention protects the rights of users by obtaining the authorization of the subjects through methods such as acquiring data, obtaining informed consent forms, and authorizing authorization forms.
[0029] In one alternative implementation, a prediction rank display module is used to display the prediction rank results of the subject's ten-year risk of sudden cardiac death;
[0030] The Risk Level Explanation and Suggestion module is used to explain the predicted risk levels displayed by the model and provide corresponding lifestyle and medical advice.
[0031] This invention provides a visual representation of the predicted risk level of a subject's ten-year risk of sudden cardiac death and offers corresponding suggestions, which can help them accurately monitor and receive earlier warnings in their daily lives.
[0032] Thirdly, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the method for predicting the risk of sudden cardiac death over ten years as described in the first aspect or any corresponding embodiment thereof.
[0033] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to perform the method for predicting the risk of sudden cardiac death within ten years as described in the first aspect or any corresponding embodiment thereof. Attached Figure Description
[0034] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0035] Figure 1 This is a flowchart illustrating the method for predicting the risk of sudden cardiac death over ten years according to an embodiment of the present invention.
[0036] Figure 2 This is a schematic diagram of generating a single image by connecting six first-minute three-lead waveforms from one hour together, as provided in an embodiment of the present invention.
[0037] Figure 3 This is a schematic diagram provided by an embodiment of the present invention, showing how a split and recombined three-dimensional tensor is input into a convolutional neural network model to obtain a prediction level result;
[0038] Figure 4 This is a structural block diagram of an embodiment of a ten-year risk prediction system for sudden cardiac death according to an embodiment of the present invention;
[0039] Figure 5 This is a structural block diagram of another embodiment of a system for predicting the risk of sudden cardiac death over ten years according to an embodiment of the present invention;
[0040] Figure 6 This is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. Detailed Implementation
[0041] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0042] According to an embodiment of the present invention, a method for predicting the risk of sudden cardiac death over ten years is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.
[0043] This embodiment provides a method for predicting the risk of sudden cardiac death over ten years, which can be used on computer devices. Figure 1 This is a flowchart of a method for predicting the ten-year risk of sudden cardiac death according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps:
[0044] Step S101: Obtain 48 hours of ECG data from a preset number of individual samples and the data required for the China-PAR risk assessment model. Generate a risk grading assessment result based on the China-PAR scale based on the data collected by the China-PAR risk assessment model, and divide it into a training set and a test set according to a preset ratio.
[0045] The data collected based on the China-PAR risk assessment model in this embodiment of the invention includes: each subject's gender, age, current residence, region, waist circumference, total cholesterol, high-density lipoprotein cholesterol, current blood pressure level, whether they are taking antihypertensive drugs, whether they have diabetes, whether they currently smoke, and whether they have a family history of cardiovascular disease.
[0046] In one specific embodiment, data from approximately 6,000 employees of a certain organization, aged 23 to 60, were collected based on the China-PAR risk assessment model, and the assessment results were calculated. The China-PAR risk assessment model generally assesses the risk level of the subjects over a ten-year period. Electrocardiogram (ECG) physiological signals (including exercise and sleep periods) were collected throughout the day using a 48-hour three-lead dynamic ECG wearable device. The scale assessment results were divided into three levels: low risk, intermediate risk, and high risk, with 200 people selected from each level, totaling 600 people. These 600 people were divided into a training set (540 people, 180 people in each level) and a test set (60 people, 20 people in each level).
[0047] Step S102 involves splitting and combining the electrocardiogram data according to different dimensions to obtain a preset multi-dimensional waveform. The specific process is as follows:
[0048] A1. Divide the 48-hour ECG data from the three-lead ECG into two 24-hour periods, then divide the 24-hour periods into 24 one-hour periods, and finally divide the one-hour periods into 6 ten-minute periods.
[0049] A2, generate a three-lead waveform for the first minute of each of the ten one-minute intervals, then connect the three-lead waveforms for the first minute of each of the six one-minute intervals within one hour to generate a combined waveform (e.g., Figure 2 (as shown in the connected graph), and finally, the connected graphs generated in each hour of the 24-hour period are superimposed to form a three-dimensional tensor;
[0050] A3. Repeat the above operation to generate a three-dimensional tensor for the second minute of each of the six 10-minute intervals, until all 10 three-dimensional tensors in 24 hours are generated. Thus, each individual's 48-hour ECG data will form 20 three-dimensional tensors.
[0051] This invention utilizes three-lead ECG data, which can be acquired using portable, high-precision ECG acquisition and recording devices, providing a comprehensive view of the subject's ECG data without burdening subsequent data processing. By breaking down the three-lead ECG data into different subdivisions and combining them, the neural network can learn the impact of ECG data at different time points on risk levels from various dimensions.
[0052] Step S103: Construct a convolutional neural network model, take the preset multi-dimensional waveforms of the training set as input, take the corresponding risk grading assessment results based on the China-PAR scale as output, train the convolutional neural network model, and verify it with a test set to obtain a sudden cardiac death risk prediction model that meets the preset requirements.
[0053] Convolutional neural networks can automatically learn features from large-scale data and generalize the results to unknown data of the same type. Their advantage is more pronounced when the input is an image, as images can be directly used as network input, avoiding the complex feature extraction process of traditional recognition algorithms. This gives them a significant advantage in two-dimensional image processing. For example... Figure 3 As shown, in this embodiment of the invention, the recombined three-dimensional tensor is input into a convolutional neural network model to learn the survival probability of subjects, as well as the corresponding causative factors, triggers, and electrocardiogram signals in the development of sudden cardiac death, thereby obtaining a prediction level result for detecting the risk of cardiac arrest in daily life outside of hospitals. In a specific embodiment, a 96% accuracy on the test set indicates a satisfactory sudden cardiac death risk prediction model.
[0054] Step S104: Obtain 48 hours of electrocardiogram data from the subject, and perform different dimensions of splitting and combination processing to obtain a preset multi-dimensional waveform. Input the waveform into the sudden cardiac death risk prediction model, and the output result is used as the prediction level of the ten-year sudden cardiac death risk.
[0055] In practical applications, the unique code marked on the portable electrocardiogram device can be scanned to establish an account association between the subject and the device, and enable the storage of the dynamic electrocardiogram data.
[0056] The present invention provides a method for predicting the risk of sudden cardiac death over ten years. This method involves creating a non-linear, high-precision, iterative deep learning model for the risk of sudden cardiac death in the general population and among physicians. This model provides a low-cost, non-invasive, and immediate prediction of the risk of sudden cardiac death (coronary angiography is currently one of the most accurate methods for diagnosing cardiovascular diseases; however, it is a cumbersome and time-consuming process requiring invasive procedures and contrast agents, which may cause certain complications such as bleeding, vascular injury, and arrhythmia). The resulting predictions are validated by physicians and matched with clinical understanding, thereby reducing diagnostic time, saving unnecessary testing time, improving the detection rate of sudden cardiac death, and reducing recurrence and mortality rates.
[0057] This embodiment also provides a ten-year predictive system for the risk of sudden cardiac death, which is used to implement the above embodiments and preferred embodiments, and will not be repeated hereafter. As used below, the term "module" can be a combination of software and / or hardware that performs a predetermined function. Although the system described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.
[0058] This embodiment provides a ten-year prediction system for the risk of sudden cardiac death, such as... Figure 4 As shown, it includes:
[0059] The sample data acquisition module 401 is used to acquire 48 hours of electrocardiogram data of a preset number of individual samples and the data required for the China-PAR risk assessment model, and to generate risk grading assessment results based on the China-PAR scale based on the data collected by the China-PAR risk assessment model, which are divided into training set and test set according to a preset ratio.
[0060] The electrocardiogram data processing module 402 is used to split and combine the electrocardiogram data in different dimensions to obtain a preset multi-dimensional waveform.
[0061] The sudden cardiac death risk prediction model acquisition module 403 is used to construct a convolutional neural network model. It takes the preset multi-dimensional waveform of the training set as input, the corresponding risk classification assessment result based on the China-PAR scale as output, trains the convolutional neural network model, and verifies it with a test set to obtain a sudden cardiac death risk prediction model that meets the preset requirements.
[0062] The ten-year cardiac sudden death risk prediction module 404 is used to acquire 48 hours of electrocardiogram data of the subject, and perform different dimensions of splitting and combination processing to obtain a preset multi-dimensional waveform diagram, which is input into the cardiac sudden death risk prediction model. The output result is used as the prediction level result of the ten-year cardiac sudden death risk.
[0063] In this embodiment, Figure 5 As shown, the prediction system described above also includes:
[0064] The user authorization module 405 is used to obtain the informed consent form of the subject. In this embodiment of the invention, the authorization of the subject is obtained by acquiring data, obtaining informed consent forms, etc., so as to protect the rights and interests of the user.
[0065] The prediction level display module 406 is used to display the prediction level results of the subject's ten-year risk of sudden cardiac death. The prediction level display module can generate health indicators in the form of 3D visualization through visual image generation. Through data statistics, it can discover the individual's own risk stratification and the individual's risk stratification in a group of people in the data matrix. For example, if the subject is female, it can display the risk level of women of similar age to the subject and the risk level of women in the same occupation.
[0066] The risk level explanation and recommendation module 407 is used to explain the predicted risk level results of the demonstration model and provide corresponding lifestyle and medical advice. Specifically, for example, if the risk level is low, it may be recommended to maintain a healthy lifestyle, maintain a balanced diet, increase the intake of whole grains and vegetables, engage in appropriate exercise that is beneficial to prevent cardiovascular disease, reduce sedentary time, maintain 6-8 hours of sleep, and regularly monitor indicators such as heart rate, blood pressure, and cholesterol.
[0067] The prediction system provided in this invention is compatible with and easily expandable to medical-grade mobile ECG monitors, hospital medical record systems, physical examination systems, and enterprise health, safety, and environmental protection systems. From signing the informed consent form to performing the ECG and obtaining the result of sudden cardiac death, the entire process is completed within 48 hours. The prediction process only takes a few seconds. The generated prediction results are verified by doctors and match clinical understanding, which can reduce diagnosis time and save unnecessary testing time.
[0068] The ten-year risk prediction system for sudden cardiac death in this embodiment is presented in the form of functional units. Here, a unit refers to an ASIC circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.
[0069] Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.
[0070] This invention also provides a computer device having the above-described features. Figure 4 or Figure 5 The system shown is a prediction system for the ten-year risk of sudden cardiac death.
[0071] Please see Figure 6 , Figure 6 This is a schematic diagram of the structure of a computer device provided in an optional embodiment of the present invention, such as... Figure 6As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on an external input / output system (such as a display device coupled to the interface). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 6 Take a processor 10 as an example.
[0072] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.
[0073] The memory 20 stores instructions executable by at least one processor 10 to cause the at least one processor 10 to perform the method shown in the above embodiments.
[0074] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device as shown by a landing page for an app. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, which can be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.
[0075] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.
[0076] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.
[0077] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.
[0078] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.
Claims
1. A method of predicting the risk of sudden cardiac death of a subject within ten years, performed by a computer, characterized in that, The method includes: Acquire 48 hours of ECG data from a preset number of individual samples and the data required for the China-PAR risk assessment model. Generate a risk grading assessment result based on the China-PAR scale based on the data collected by the China-PAR risk assessment model. Divide the data into a training set and a test set according to a preset ratio. The ECG data is three-lead ECG data. The electrocardiogram (ECG) data is split and combined in different dimensions to obtain a preset multi-dimensional waveform. This includes: dividing the 48-hour ECG data of the three-lead ECG into two 24-hour periods, dividing each 24-hour period into 24 one-hour periods, and dividing each one-hour period into six 10-minute periods; generating a three-lead waveform for the first minute of each of the six 10-minute periods; connecting the three-lead waveforms of the first minute of each of the six one-minute periods in one hour to generate a composite image; and finally, stacking the composite images generated in each hour of the 24-hour period together to form a three-dimensional tensor; repeating the above operations to generate a three-dimensional tensor for the second minute of each of the six 10-minute periods until 10 three-dimensional tensors in the 24-hour period are generated. Thus, each individual's 48-hour ECG data forms 20 three-dimensional tensors. A convolutional neural network model is constructed, taking the pre-defined multi-dimensional waveforms of the training set as input and the corresponding risk grading assessment results based on the China-PAR scale as output. The convolutional neural network model is trained and validated using a test set to obtain a sudden cardiac death risk prediction model that meets the pre-defined requirements. The subject's 48-hour electrocardiogram data was acquired and processed by splitting and combining data in different dimensions to obtain a preset multi-dimensional waveform. This waveform was then input into the sudden cardiac death risk prediction model, and the output results served as the prediction level of the ten-year sudden cardiac death risk.
2. The method of claim 1, wherein, Data collected based on the China-PAR risk assessment model includes: each participant's gender, age, current residence, region, waist circumference, total cholesterol, high-density lipoprotein cholesterol, current blood pressure level, whether they are taking antihypertensive medication, whether they have diabetes, whether they currently smoke, and whether they have a family history of cardiovascular disease.
3. The method of claim 1, wherein, Before obtaining the subject's 48-hour electrocardiogram data, the process also includes obtaining the subject's informed consent form.
4. A ten-year predictive system for the risk of sudden cardiac death, characterized in that, include: The sample data acquisition module is used to acquire 48-hour electrocardiogram data of a preset number of individual samples and the data required for the China-PAR risk assessment model. Based on the data collected by the China-PAR risk assessment model, it generates a risk grading assessment result based on the China-PAR scale and divides it into a training set and a test set according to a preset ratio. The electrocardiogram data is three-lead electrocardiogram data. The electrocardiogram (ECG) data processing module is used to split and combine the ECG data in different dimensions to obtain preset multi-dimensional waveforms. This includes: dividing the 48-hour ECG data into two 24-hour periods; dividing each 24-hour period into 24 one-hour periods; and dividing each one-hour period into six 10-minute periods. A three-lead waveform is generated for the first minute of each of the six 10-minute periods. The six first-minute three-lead waveforms within each hour are then connected together to generate a composite image. Finally, the composite images generated for each hour within the 24-hour period are overlaid to form a three-dimensional tensor. This process is repeated to generate a three-dimensional tensor for the second minute of each of the six 10-minute periods, until all 10 three-dimensional tensors for the 24-hour period are generated. Thus, each individual's 48-hour ECG data generates 20 three-dimensional tensors. The module for obtaining a risk prediction model for sudden cardiac death is used to construct a convolutional neural network model. It takes a preset multi-dimensional waveform of the training set as input, the corresponding risk grading assessment result based on the China-PAR scale as output, trains the convolutional neural network model, and verifies it with a test set to obtain a risk prediction model for sudden cardiac death that meets the preset requirements. The ten-year cardiac sudden death risk prediction module is used to acquire 48 hours of electrocardiogram data from the subject, and to perform different dimensions of splitting and combination processing to obtain a preset multi-dimensional waveform. This waveform is then input into the cardiac sudden death risk prediction model, and the output result serves as the prediction level of the ten-year cardiac sudden death risk.
5. The system of claim 4, wherein, Also includes: The user authorization module is used to obtain informed consent from the subjects.
6. The system of claim 5, wherein, Also includes: The prediction level display module is used to display the prediction level results of the subject's ten-year risk of sudden cardiac death; The Risk Level Explanation and Suggestion module is used to explain the predicted risk levels displayed by the model and provide corresponding lifestyle and medical advice.
7. A computer device, comprising: include: A memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, the processor executing the computer instructions to perform the computer-executed method for predicting the risk of sudden cardiac death over ten years as described in any one of claims 1 to 3.
8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to perform the computer-executed method for predicting the risk of sudden cardiac death over ten years, as described in any one of claims 1 to 3.