Cardiovascular disease risk prediction method and system

Abstract

The invention relates to the technical field of medical health, and discloses a cardiovascular disease risk prediction method and system, and the method comprises the steps: obtaining the physiological parameter data of a user in real time, obtaining the life behavior data of the user in real time, extracting a real-time feature vector from the preprocessed real-time physiological parameter data and real-time life behavior data, and carrying out the real-time feature vector prediction. The real-time feature vector comprises a plurality of real-time feature factors and the factor weight of each real-time feature factor, screening the real-time feature factors from the real-time feature vector according to the factor weights, inputting the real-time feature factors into a pre-trained risk prediction model, and outputting a cardiovascular disease prediction result. Physiological parameter data and life behavior data of the body health state of a user are comprehensively reflected, factor weights are given to real-time features of real-time feature vectors based on the correlation between parameters and risks and the correlation between the parameters, and feature factors with higher factor weights are selected to predict risks. The generalization ability and precision of the prediction result are improved, and a real-time risk assessment result is obtained.

Description

Technical Field

[0001] The present invention relates to the field of medical health technology, and in particular to a method and system for predicting cardiovascular disease risk. Background Art

[0002] Cardiovascular diseases (CVDs) are one of the leading causes of death worldwide, causing millions of deaths each year. Since the early symptoms of cardiovascular diseases are often not obvious, many patients are not diagnosed until their condition becomes serious, which greatly reduces the treatment effect and survival rate. Therefore, the development of effective cardiovascular disease risk prediction methods is of great significance for early identification of high-risk groups and implementation of preventive measures. Currently, there are a variety of cardiovascular disease risk prediction tools and methods, including the Framingham heart disease risk score, ASCVD risk assessment, QRISK® model and SCORE model. These models mainly rely on static demographic and clinical indicators, such as age, gender, blood lipid levels, blood pressure and smoking status, and fail to fully reflect the health status and dynamic changes of individuals. However, the existing cardiovascular disease risk prediction methods have the following major problems: first, the prediction accuracy is limited because these models mainly rely on traditional medical examination data, such as blood pressure, blood lipid levels, etc., while ignoring other important physiological and psychological indicators; second, there is a lack of personalization. There are significant differences in genetic background, lifestyle and environmental factors among different individuals. These differences have an important impact on the risk of cardiovascular disease. Existing models often fail to fully consider these individual differences, resulting in insufficient generalization of prediction results; finally, the dynamic monitoring capability is poor. Existing risk prediction models mainly rely on one-time static data collection, such as annual physical examination data. This method cannot reflect the dynamic changes in individual health status in a timely manner, resulting in a lag in risk assessment results. In addition, there is a lack of effective continuous monitoring methods and it is impossible to provide real-time risk assessment results. Summary of the invention

[0003] In order to solve the above technical problems, the present invention provides a method for predicting cardiovascular disease risk, comprising: Step 1, obtaining the user's physiological parameter data in real time, wherein the physiological parameter data includes blood pressure, blood lipid level, blood sugar level, and heart rate variability; Step 2, acquiring the user's life behavior data in real time, wherein the life behavior data includes eating habits, exercise frequency, sleep quality, and smoking status; Step 3, preprocessing the real-time physiological parameter data and real-time life behavior data; Step 4, extracting a real-time feature vector from the real-time physiological parameter data and the real-time life behavior data after preprocessing; the real-time feature vector includes a number of real-time feature factors and a factor weight of each real-time feature factor; the real-time feature vector includes one or more real-time physiological parameter data, and / or, one or more real-time life behavior data; Step 5, screen the real-time feature factors from the real-time feature vector according to the factor weights, input them into the pre-trained risk prediction model, and output the cardiovascular disease prediction results.

[0004] Furthermore, the real-time feature vector also includes additional features, and the real-time additional feature vector is obtained, which specifically includes: Step 41, using a sliding window technique to perform time series analysis on the real-time physiological parameter data to extract real-time feature values; Step 42, using a clustering algorithm to perform cluster analysis on the real-time life behavior data to identify real-time user behavior patterns; Step 43: Add the real-time feature value and the real-time user behavior pattern as additional features to the real-time feature vector to obtain a real-time additional feature vector.

[0005] Furthermore, the factor weight of each real-time feature factor is obtained, including: Step 44, obtaining a first risk correlation between each real-time psychological parameter and cardiovascular risk; Step 45, obtaining a second risk correlation between each real-time life behavior data and cardiovascular risk; Step 46, obtaining the parameter correlation between each real-time psychological parameter and each real-time life behavior data; Step 47, obtaining a correction value according to the ratio among the first risk correlation, the second risk correlation and the parameter correlation; Step 48, obtaining factor weights according to the first risk correlation, the second risk correlation, the parameter correlation, and the correction value.

[0006] Furthermore, a prediction function is constructed according to the factor weights and the characteristic values ​​of the characteristic factors selected according to the factor weights to obtain a prediction result of cardiovascular disease.

[0007] Furthermore, the first risk correlation, the second risk correlation and the parameter correlation are calculated by Pearson's correlation coefficient.

[0008] Furthermore, the correction value is: ; In the formula, represents the correction value, represents parameter correlation, represents the first risk correlation, Represents the second risk correlation.

[0009] Furthermore, the factor weights are: ; In the formula, represents the factor weight of the f-th characteristic factor, represents the first adjustment coefficient, , represents the parameter correlation between the fth characteristic factor and the gth characteristic factor, n represents the maximum number of characteristic factors, represents the second adjustment factor, , represents the first risk correlation between the fth eigenfactor and cardiovascular risk, represents the second risk correlation between the fth eigenfactor and cardiovascular risk, represents the correction value, represents the third adjustment factor, .

[0010] Furthermore, the prediction function is: ; In the formula, represents the prediction result of cardiovascular disease, m represents the intercept term, represents the factor weight of the g-th feature factor, Represents the eigenvalue of the g-th eigenfactor.

[0011] The present invention also provides a cardiovascular disease risk prediction system, comprising the following modules: Physiological parameter data real-time acquisition module: used to obtain the user's physiological parameter data in real time; Life behavior data real-time acquisition module: connected to the physiological parameter data real-time acquisition module, used to obtain the user's life behavior data in real time; Preprocessing module: connected with the physiological parameter data real-time acquisition module and the life behavior data real-time acquisition module, used for preprocessing the real-time physiological parameter data and the real-time life behavior data; Real-time feature vector extraction module: connected with the preprocessing module, used to extract real-time feature vectors from real-time physiological parameter data and real-time life behavior data after preprocessing; Prediction module: connected to the real-time feature vector extraction module, used to screen feature factors from the feature vector according to factor weights, input the pre-trained risk prediction model, and output the cardiovascular disease prediction results.

[0012] The embodiments of the present invention have the following technical effects: The present invention not only considers physiological parameter data to realize cardiovascular disease risk prediction, but also considers life behavior data, and both physiological parameter data and life behavior data are acquired in real time, and further obtains real-time feature vectors, and finally selects features highly correlated with cardiovascular risk from real-time feature vectors based on factor weights to realize prediction. The present invention considers physiological parameter data and life behavior data that can fully reflect the user's physical health status, and assigns factor weights to the real-time features of the real-time feature vector based on the correlation between parameters and risks and the correlation between parameters, selects feature factors with higher factor weights to predict risks, improves the generalization ability and accuracy of the prediction results, and obtains real-time risk assessment results.

[0013] The real-time feature vector and real-time feature factor of the present invention can timely reflect the dynamic changes of the user's individual health status, avoid the lag of risk assessment results, realize effective continuous monitoring means, and thus provide real-time risk assessment results.

[0014] The present invention obtains a correction value by: obtaining a first risk correlation between each real-time psychological parameter and cardiovascular risk, a second risk correlation between each real-time life behavior data and cardiovascular risk, and a parameter correlation between each real-time psychological parameter and each real-time life behavior data, and finally obtaining a factor weight according to the correction value, the first risk correlation, the second risk correlation, and the parameter correlation. The correction value quantitatively evaluates the interaction between physiological parameter data and life behavior data, reflects the relative importance of physiological parameter data and life behavior data in the prediction of cardiovascular disease risk, helps the model to more accurately capture the complex interaction between different features, and improves the flexibility and adaptability of the model. The factor weight enables the model to more accurately select the features that contribute most to the prediction of cardiovascular disease risk, and improves the prediction accuracy and personalization ability of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 is a flow chart of a cardiovascular disease risk prediction method provided by an embodiment of the present invention; Figure 2 is a schematic diagram of the structure of a cardiovascular disease risk prediction system provided by an embodiment of the present invention; Figure 3It is a structural schematic diagram of an electronic device provided by the implementation of the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work belong to the scope of protection of the present invention.

[0018] Figure 1 is a flow chart of a cardiovascular disease risk prediction method provided by an embodiment of the present invention. Figure 1 , including: Step 1, obtaining the user's physiological parameter data in real time, wherein the physiological parameter data includes blood pressure, blood lipid level, blood sugar level, and heart rate variability.

[0019] Blood pressure monitor: This is used to measure the user's blood pressure. You can choose an automatic electronic blood pressure monitor, which usually has Bluetooth or Wi-Fi connectivity and can transmit data in real time to a smartphone or other device.

[0020] Blood lipid tester: used to measure the user's blood lipid level. Portable blood lipid testers can be used to perform rapid tests by drawing blood from the fingertips. These devices usually also support wireless data transmission.

[0021] Blood glucose meter: used to measure the user's blood glucose level. Portable blood glucose meters measure blood through fingertips and also support wireless data transmission.

[0022] Heart rate monitor: used to monitor the user's heart rate variability. You can use a smart bracelet or smart watch, which usually has a heart rate monitoring function and can transmit data in real time.

[0023] The above-mentioned blood pressure monitors, blood lipid testers, blood glucose meters, and heart rate monitors are all connected to the health monitoring application, and the measurement results are automatically transmitted to the health monitoring application via Bluetooth or Wi-Fi. If abnormal data is detected, the application will immediately issue an alarm and provide corresponding suggestions, such as suggesting that users seek medical treatment in time. Users can record abnormal conditions and symptoms for doctors to diagnose and treat.

[0024] Step 2, acquiring the user's life behavior data in real time, wherein the life behavior data includes eating habits, exercise frequency, sleep quality, and smoking status.

[0025] Users can use the diet record function in the health monitoring application to record the intake of three meals and snacks every day, including food type, quantity, time and nutritional content.

[0026] Smart bracelets / smart watches automatically monitor the user's steps, exercise time and intensity.

[0027] Smart bracelets / smart watches automatically monitor the user's sleep time, deep sleep and light sleep time, number of tossing and turning, etc.

[0028] Users can use the questionnaire function in the health monitoring application to regularly record their smoking habits, including the number of times they smoke each day and the time of each smoking session.

[0029] All collected life behavior data is stored in real time in the health monitoring application on the smartphone. The application regularly synchronizes the data to the cloud server for subsequent data analysis and risk prediction. If abnormal data is detected, the application will immediately issue an alarm and provide corresponding suggestions, such as suggesting that users adjust their diet structure, increase exercise or consult a doctor.

[0030] Step 3: preprocess the real-time physiological parameter data and real-time life behavior data.

[0031] Data Collection Physiological parameter data: blood pressure, blood lipid levels, blood sugar levels, heart rate variability.

[0032] Life behavior data: eating habits, exercise frequency, sleep quality, smoking status.

[0033] Data cleaning Remove invalid data: Check whether the data is reasonable, such as whether the blood pressure value is within the normal range, and whether the blood lipid and blood sugar values ​​are positive.

[0034] Handling missing data: For missing data, you can use interpolation methods (such as linear interpolation) or use the mean value to fill it.

[0035] Detect outliers: Use statistical methods such as Z-score or IQR to detect and handle outliers.

[0036] Data Standardization Blood pressure: The unit is mmHg, no conversion is required.

[0037] Blood lipid level: The unit is mg / dL, no conversion is required.

[0038] Blood glucose level: The unit is mg / dL, no conversion is required.

[0039] Heart rate variability: The unit is milliseconds (ms), no conversion is required.

[0040] Eating habits: Convert food types and amounts into nutritional components (e.g., calories, protein, fat, carbohydrates).

[0041] Exercise frequency: Convert exercise time and intensity into standard units (e.g., minutes / day, METs).

[0042] Sleep quality: Convert sleep time and quality indicators (such as deep sleep time, light sleep time, and number of tossing and turning) into standard units (such as hours and minutes).

[0043] Smoking status: Convert the number and duration of smoking into standard units (e.g., cigarettes / day, hours / day).

[0044] Data Normalization Use Min-Max Scaling to scale all data to the range of 0 to 1.

[0045] Data extraction Physiological parameter data: Extract systolic and diastolic blood pressure.

[0046] Blood lipid levels of total cholesterol, high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), and triglycerides (TG) were extracted.

[0047] Fasting blood glucose and postprandial blood glucose levels were extracted.

[0048] Extract indicators such as SDNN (standard deviation) and RMSSD (square root of the difference between adjacent intervals) of heart rate variability.

[0049] Life behavior data: Extract total calories, protein, fat, carbohydrates, etc. from your eating habits.

[0050] Extract the total exercise time, high-intensity exercise time, moderate-intensity exercise time, etc. from the exercise frequency.

[0051] Extract the total sleep time, deep sleep time, light sleep time, number of tossing and turning, etc. from the sleep quality.

[0052] Extract the number of times smoked per day, smoking duration, etc. from the smoking status.

[0053] Data Integration The preprocessed physiological parameter data and life behavior data are integrated together to form a complete data set. For example, all the data can be integrated into a table, where each row represents the data at a time point and each column represents a feature.

[0054] Step 4, extracting a real-time feature vector from the real-time physiological parameter data and real-time life behavior data after preprocessing; the real-time feature vector includes several real-time feature factors and factor weights of each real-time feature factor; the real-time feature vector includes one or several real-time physiological parameter data, and / or, one or several real-time life behavior data.

[0055] The real-time feature vector also includes additional features, and the real-time additional feature vector is obtained, which specifically includes: Step 41, using sliding window technology to perform time series analysis on the real-time physiological parameter data to extract real-time feature values.

[0056] The sliding window setting includes a window size and a sliding step size. For example, the window size is 30 minutes or 1 hour, and the sliding step size is 10 minutes or 15 minutes.

[0057] Time series data extraction: Blood pressure: average systolic and diastolic blood pressure within the extraction window. Blood lipid level: average total cholesterol, HDL-C, LDL-C and triglycerides within the extraction window. Blood glucose level: average fasting blood glucose and postprandial blood glucose within the extraction window. Heart rate variability: average SDNN and RMSSD within the extraction window.

[0058] Eigenvalue calculation: Statistical features: Calculate the mean, standard deviation, maximum, minimum and other statistical features of the data in the window. Frequency domain features: Calculate the spectrum features of the data in the window, such as power spectral density (PSD). Time features: Calculate the time features of the data in the window, such as the rate of change and fluctuation amplitude.

[0059] Step 42: Perform cluster analysis on the real-time life behavior data using a clustering algorithm to identify real-time user behavior patterns.

[0060] Select features that have a greater impact on user behavior patterns, such as total calories, total exercise time, total sleep time, and number of cigarettes smoked per day. Select an appropriate number of clusters (such as 3 or 4), use the elbow rule to determine the best K value, train the K-means model, and assign user behavior data to different clusters, each cluster representing a user behavior pattern.

[0061] Step 43: Add the real-time feature value and the real-time user behavior pattern as additional features to the real-time feature vector to obtain a real-time additional feature vector.

[0062] The physiological parameter feature values ​​(such as mean systolic blood pressure, standard deviation, etc.) extracted by the sliding window are added to the feature vector, and the user behavior pattern labels obtained by cluster analysis are added to the feature vector. All features are integrated into a complete feature vector for subsequent analysis and prediction.

[0063] Obtaining the factor weight of each real-time feature factor, including: Step 44, obtaining a first risk correlation between each real-time psychological parameter data and cardiovascular risk.

[0064] Step 45, obtaining the second risk correlation between each real-time life behavior data and cardiovascular risk.

[0065] Step 46, obtaining the parameter correlation between each real-time psychological parameter and each real-time life behavior data.

[0066] The cardiovascular risk is calculated by risk scoring. Preferably, the risk score is between 0 and 1, the closer to 1, the higher the cardiovascular risk, and the closer to 0, the lower the cardiovascular risk.

[0067] The first risk correlation, the second risk correlation and the parameter correlation are calculated by Pearson's correlation coefficient.

[0068] Step 47, obtaining a correction value according to the ratio among the first risk correlation, the second risk correlation and the parameter correlation.

[0069] The pathogenesis of cardiovascular disease is the result of the combined effects of multiple factors. Simply considering physiological parameters or life behavior data may not fully reflect the risk. By introducing correction values, the interaction between multiple factors can be comprehensively considered.

[0070] If certain physiological parameters or life behavior data vary significantly between different users, or if physiological parameters or life behavior data fluctuate, the correction value can help the model better balance these differences or fluctuations and ensure that the weight of each feature in risk prediction is more reasonable.

[0071] The correction value can also capture the relationship between physiological parameter data and life behavior data. The correction value can balance the impact of physiological parameter data and life behavior data on cardiovascular risk. If there is a high correlation between a physiological parameter and a life behavior data, the correction value will increase the weight of this combination to better reflect its impact on cardiovascular risk. If a user's blood pressure suddenly rises, but his exercise frequency and sleep quality are good, the correction value can help the model comprehensively consider these factors and avoid misjudging the user's cardiovascular risk.

[0072] ; In the formula, represents the correction value, represents parameter correlation, represents the first risk correlation, Represents the second risk correlation.

[0073] Step 48, obtaining factor weights according to the first risk correlation, the second risk correlation, the parameter correlation, and the correction value.

[0074] ; In the formula, represents the factor weight of the f-th characteristic factor, represents the first adjustment coefficient, , represents the parameter correlation between the fth characteristic factor and the gth characteristic factor, n represents the maximum number of characteristic factors, represents the second adjustment factor, , represents the first risk correlation between the fth eigenfactor and cardiovascular risk, represents the second risk correlation between the fth eigenfactor and cardiovascular risk, represents the correction value, represents the third adjustment factor, .

[0075] In this embodiment, the first adjustment coefficient Give: , which means that the correlation between parameters is considered an important factor and plays a leading role in determining factor weights. This is done to emphasize the interrelationships between different feature factors, because they work together to contribute to the final risk assessment results. By giving higher weights, the model pays more attention to those feature combinations with significant interaction effects, which helps to discover complex patterns hidden in the data and may reveal additional information that a single feature cannot provide alone. On the other hand, the second adjustment coefficient Give: , although these direct connections are also important, in some cases, they are not as critical or reliable as the relationship between features. For example, when there is a strong positive correlation between two variables, changes in one variable often lead to corresponding changes in the other variable; however, this simple one-to-one correspondence does not always hold true. Sometimes, even if there is no obvious causal link connecting the two, they still affect each other to some extent. Therefore, reducing the proportion of this contribution allows the model to rely more on verified and effective evidence rather than just making judgments based on superficial phenomena.

[0076] In addition, allocating weights in this way can also prevent overfitting problems. If too much emphasis is placed on any one type of information source, it may cause the model to rely too much on a specific type of input signal and ignore other potentially useful data sources. In contrast, adopting a balanced strategy can ensure that all available resources are fully utilized while maintaining overall performance. In short, this design aims to build a powerful and robust prediction system that can effectively utilize the existing knowledge base and flexibly respond to new challenges that may arise in the future.

[0077] Step 5, screen the real-time feature factors from the real-time feature vector according to the factor weights, input them into the pre-trained risk prediction model, and output the cardiovascular disease prediction results.

[0078] A prediction function is constructed according to the factor weights and the characteristic values ​​of the characteristic factors selected according to the factor weights to obtain the prediction results of cardiovascular diseases.

[0079] The prediction function is: ; In the formula, represents the prediction result of cardiovascular disease, m represents the intercept term, and m is a constant used to adjust the overall offset of the prediction function. represents the factor weight of the g-th feature factor, Represents the eigenvalue of the g-th eigenfactor.

[0080] The prediction result calculated by the above formula is a value between 0 and 1, which indicates the risk of cardiovascular disease. The closer the value is to 1, the higher the risk of cardiovascular disease, and the closer the value is to 0, the lower the risk of cardiovascular disease. When the risk level predicted by the user reaches about 0.6, in order to avoid sudden cardiovascular disease in the future, you should seek medical attention in time for a comprehensive examination to achieve a preventive effect.

[0081] like Figure 2 As shown, this embodiment also provides a cardiovascular disease risk prediction system, including the following modules: Physiological parameter data real-time acquisition module: used to obtain the user's physiological parameter data in real time; Life behavior data real-time acquisition module: connected to the physiological parameter data real-time acquisition module, used to obtain the user's life behavior data in real time; Preprocessing module: connected with the physiological parameter data real-time acquisition module and the life behavior data real-time acquisition module, used for preprocessing the real-time physiological parameter data and the real-time life behavior data; Real-time feature vector extraction module: connected with the preprocessing module, used to extract real-time feature vectors from real-time physiological parameter data and real-time life behavior data after preprocessing; Prediction module: connected to the real-time feature vector extraction module, used to screen feature factors from the feature vector according to factor weights, input the pre-trained risk prediction model, and output the cardiovascular disease prediction results.

[0082] Figure 3 Schematic diagram of the structure of an electronic device provided by the present invention. Figure 3 As shown, the electronic device 500 includes one or more processors 501 and a memory 502 .

[0083] The processor 501 may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 500 to perform desired functions.

[0084] The memory 502 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory (cache), etc. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 501 may run the program instructions to implement a cardiovascular disease risk prediction method and / or other desired functions of any embodiment of the present application described above. Various contents such as initial external parameters, thresholds, etc. may also be stored in the computer-readable storage medium.

[0085] In one example, the electronic device 500 may further include: an input device 503 and an output device 504, which are interconnected via a bus system and / or other forms of connection mechanisms (not shown). The input device 503 may include, for example, a keyboard, a mouse, etc. The output device 504 may output various information to the outside, including early warning prompt information, braking force, etc. The output device 504 may include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, etc.

[0086] Of course, to simplify, Figure 3 Only some of the components related to the present application in the electronic device 500 are shown, and components such as a bus, an input / output interface, etc. are omitted. In addition, according to specific application conditions, the electronic device 500 may also include any other appropriate components.

[0087] In addition to the above-mentioned methods and devices, an embodiment of the present application may also be a computer program product, which includes computer program instructions, which, when executed by a processor, enable the processor to execute the steps of a cardiovascular disease risk prediction method provided in any embodiment of the present application.

[0088] The computer program product may be written in any combination of one or more programming languages ​​to write program codes for performing the operations of the embodiments of the present application, including object-oriented programming languages, such as Java, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as an independent software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0089] In addition, an embodiment of the present application may also be a computer-readable storage medium on which computer program instructions are stored. When the computer program instructions are executed by a processor, the processor executes the steps of a cardiovascular disease risk prediction method provided in any embodiment of the present application.

[0090] The computer readable storage medium can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can include, for example, but is not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0091] It should be noted that the terms used in the present invention are only for describing specific embodiments, rather than limiting the scope of the present application. As shown in the present specification, unless the context clearly indicates an exception, the words "one", "a", "a kind of" and / or "the" do not specifically refer to the singular, but may also include the plural. The terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method or device. In the absence of more restrictions, the elements defined by the sentence "include one..." do not exclude the presence of other identical elements in the process, method or device including the elements.

[0092] It should also be noted that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. Unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", etc. should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be a connection between the two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0093] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting cardiovascular disease risk, characterized in that: include: Step 1, obtaining the user's physiological parameter data in real time, wherein the physiological parameter data includes blood pressure, blood lipid level, blood sugar level, and heart rate variability; Step 2, acquiring the user's life behavior data in real time, wherein the life behavior data includes eating habits, exercise frequency, sleep quality, and smoking status; Step 3, preprocessing the real-time physiological parameter data and real-time life behavior data; Step 4, extracting a real-time feature vector from the real-time physiological parameter data and the real-time life behavior data after preprocessing; the real-time feature vector includes a number of real-time feature factors and a factor weight of each real-time feature factor; the real-time feature vector includes one or more real-time physiological parameter data, and / or, one or more real-time life behavior data; Step 5, screen the real-time feature factors from the real-time feature vector according to the factor weights, input them into the pre-trained risk prediction model, and output the cardiovascular disease prediction results.

2. A cardiovascular disease risk prediction method according to claim 1, characterized in that: The real-time feature vector also includes additional features, and the real-time additional feature vector is obtained, which specifically includes: Step 41, using a sliding window technique to perform time series analysis on the real-time physiological parameter data to extract real-time feature values; Step 42, using a clustering algorithm to perform cluster analysis on the real-time life behavior data to identify real-time user behavior patterns; Step 43: Add the real-time feature value and the real-time user behavior pattern as additional features to the real-time feature vector to obtain a real-time additional feature vector.

3. A cardiovascular disease risk prediction method according to claim 2, characterized in that: Obtaining the factor weight of each real-time feature factor, including: Step 44, obtaining a first risk correlation between each real-time psychological parameter and cardiovascular risk; Step 45, obtaining a second risk correlation between each real-time life behavior data and cardiovascular risk; Step 46, obtaining the parameter correlation between each real-time psychological parameter and each real-time life behavior data; Step 47, obtaining a correction value according to the ratio among the first risk correlation, the second risk correlation and the parameter correlation; Step 48, obtaining factor weights according to the first risk correlation, the second risk correlation, the parameter correlation, and the correction value.

4. A cardiovascular disease risk prediction method according to claim 3, characterized in that: A prediction function is constructed according to the factor weights and the characteristic values ​​of the characteristic factors selected according to the factor weights to obtain the prediction results of cardiovascular diseases.

5. A cardiovascular disease risk prediction method according to claim 4, characterized in that: The first risk correlation, the second risk correlation and the parameter correlation are calculated by Pearson's correlation coefficient.

6. A cardiovascular disease risk prediction method according to claim 5, characterized in that: The correction value is: ； In the formula, represents the correction value, represents parameter correlation, represents the first risk correlation, Represents the second risk correlation.

7. A method for predicting cardiovascular disease risk according to claim 6, characterized in that: The factor weights are: ； In the formula, represents the factor weight of the f-th characteristic factor, represents the first adjustment coefficient, , represents the parameter correlation between the fth characteristic factor and the gth characteristic factor, n represents the maximum number of characteristic factors, represents the second adjustment factor, , represents the first risk correlation between the fth eigenfactor and cardiovascular risk, represents the second risk correlation between the fth eigenfactor and cardiovascular risk, represents the correction value, represents the third adjustment factor, .

8. A method for predicting cardiovascular disease risk according to claim 4, characterized in that: The prediction function is: ； In the formula, represents the prediction result of cardiovascular disease, m represents the intercept term, represents the factor weight of the g-th feature factor, Represents the eigenvalue of the g-th eigenfactor.

9. A cardiovascular disease risk prediction system, used to execute a cardiovascular disease risk prediction method according to any one of claims 1 to 8, characterized in that: Includes the following modules: Physiological parameter data real-time acquisition module: used to obtain the user's physiological parameter data in real time; Life behavior data real-time acquisition module: connected to the physiological parameter data real-time acquisition module, used to obtain the user's life behavior data in real time; Preprocessing module: connected with the physiological parameter data real-time acquisition module and the life behavior data real-time acquisition module, used for preprocessing the real-time physiological parameter data and the real-time life behavior data; Real-time feature vector extraction module: connected with the preprocessing module, used to extract real-time feature vectors from real-time physiological parameter data and real-time life behavior data after preprocessing; Prediction module: connected to the real-time feature vector extraction module, used to screen feature factors from the feature vector according to factor weights, input the pre-trained risk prediction model, and output the cardiovascular disease prediction results.

Images