A method and system for predicting the risk of coronary heart disease

Abstract

The present application relates to the field of home medical monitoring and provides a method and system for predicting the risk of coronary heart disease. The method includes obtaining behavioral data of a user after a physical examination and determining the user's behavior pattern based on the behavioral data; obtaining corresponding influence weights and influence factors based on real-time detection data under the user's behavior pattern; adjusting the most recent physical examination data based on the influence factors and influence weights to obtain predicted physical data; and initiating an alarm when the predicted physical data meets a preset health risk condition. The system fine-tunes the most recent physical examination data based on the user's behavior, thereby estimating the physical condition under the state corresponding to the current user's behavior data, and then estimating the predicted physical data to determine whether there will be health risks, thereby estimating the patient's physical condition after discharge, and making an early prediction of the physical condition.

Description

Technical Field

[0001] The present application relates to the field of intelligent healthcare, and in particular to a method and system for predicting the risk of coronary heart disease. Background Art

[0002] Coronary heart disease has a long development cycle and can be prevented and treated if detected early. Therefore, there is an urgent need for reliable coronary heart disease screening tools that can be used in the community to promote early detection, diagnosis and treatment of the disease.

[0003] Existing coronary heart disease risk prediction methods are often based on actual clinical cases, monitoring patients' current risk at a moment's notice, but are unable to track patients' daily behaviors. Furthermore, patients' families are unable to constantly monitor the disease risks associated with their behaviors. Therefore, a method and system is urgently needed that can provide comprehensive medical monitoring and predict disease risk for patients during their home care after a physical examination. Summary of the Invention

[0004] In order to solve the problem in the prior art that it is difficult to predict the risk of coronary heart disease caused by the patient's behavior after leaving the hospital, the present application provides a method and system for predicting the risk of coronary heart disease.

[0005] This application provides a method and system for predicting the risk of coronary heart disease, which adopts the following technical solutions:

[0006] A method for predicting the risk of coronary heart disease, comprising:

[0007] Obtaining user behavior data after the physical test, and determining the user's behavior pattern based on the behavior data;

[0008] Based on the real-time detection data under the user behavior pattern, the corresponding influence weight and influence factor are obtained;

[0009] Adjust the most recent physical measurement data according to the influencing factors and the influencing weights to obtain predicted physical data;

[0010] When the predicted physical data meets a preset health risk condition, an alarm is activated.

[0011] Optionally, obtain user behavior data after the physical test, including:

[0012] Periodically collect behavioral data and behavior time;

[0013] Extract all behavioral time and behavioral data after the most recent physical test time to form a first sequence set;

[0014] Extracting behavior time and behavior data whose difference from the current time in the first sequence set is within the first time range to form a second sequence set;

[0015] When the second sequence set is continuous, the behavior data in the second sequence set is used as the behavior data after the physical test.

[0016] Optionally, obtaining user behavior data after the physical test also includes:

[0017] When the second sequence set is incoherent, reading the behavior time and behavior data within a second time range before and after the behavior time corresponding to the blank behavior data, and forming a third sequence set and a fourth sequence set respectively;

[0018] According to the trends of the third sequence set and the fourth sequence set, theoretical data of the blank behavior data is predicted, and the blank behavior data is updated using the theoretical data, thereby completing the second sequence set and making it coherent.

[0019] Optionally, obtaining user behavior data after the physical test also includes:

[0020] When the second sequence set is incoherent, calculating the ratio of blank behavior data in the second sequence set to all data in the second sequence set to obtain a missing ratio;

[0021] When the missing ratio exceeds a preset effective range, a first monitoring alarm signal is generated;

[0022] When the missing ratio is within a preset effective range, blank behavior data is identified according to the behavior data in the second sequence set.

[0023] Optionally, determining a user behavior pattern based on the behavior data includes:

[0024] When the behavior data satisfies the sleep pattern, setting the user behavior mode to sleep;

[0025] When the behavior data meets the rules of vigorous exercise, the user behavior mode is set to vigorous exercise;

[0026] When the behavior pattern meets the regular exercise rules, the user behavior pattern is set as regular exercise.

[0027] Optionally, based on real-time detection data under user behavior patterns, corresponding influence weights and influence factors are derived, including:

[0028] Obtain corresponding data weight rules based on user behavior patterns;

[0029] According to each influencing factor in the real-time detection data and the data weight rule, the corresponding influence weight is obtained.

[0030] Optionally, adjusting the most recent measured physical data according to the influencing factors and the influencing weights to obtain the predicted physical data includes:

[0031] Substituting the impact weight into the initial impact model corresponding to the impact factor to obtain an existing impact model;

[0032] Using the existing influence model to adjust the physical data of the physical test to obtain individual physical data corresponding to the influence factor;

[0033] The predicted physical data is obtained comprehensively based on the individual physical data corresponding to all the influencing factors.

[0034] Optionally, when the predicted physical data meets a preset health risk condition, an alarm is initiated, including:

[0035] If any one of the predicted physical data is within a preset risk range, the predicted physical data satisfies the health risk.

[0036] Optionally, predicting theoretical data of the blank behavior data according to the trends of the third sequence set and the fourth sequence set includes:

[0037] fitting the data of the second sequence set and the fourth sequence set to obtain an initial trend model;

[0038] Screening out discrete data in the third and fourth sequence sets whose distance from the initial strike reaches a warning value, updating the second and fourth sequence sets, and re-estimating an intermediate strike model based on the updated third and fourth sequence sets;

[0039] Fitting the initial trend model and the intermediate trend model to obtain a final trend model;

[0040] Theoretical data corresponding to the blank behavior data is calculated according to the final trend model.

[0041] The present invention also provides a system for predicting the risk of coronary heart disease, the system comprising:

[0042] Information collection module, used to obtain the user's behavior data and behavior time after the physical test;

[0043] A behavior pattern determination module, configured to determine a user's behavior pattern based on the behavior data;

[0044] A body data deduction module is used to derive corresponding influence weights and influence factors based on real-time detection data under user behavior patterns, and adjust the most recent measured body data based on the influence factors and influence weights to obtain predicted body data;

[0045] The health judgment module is used to determine whether a preset health risk condition is met based on the predicted physical data, and to initiate an alarm when the predicted physical data meets the preset health risk condition.

[0046] Beneficial effects: The current behavior mode of the user is determined based on the user's behavior data, and the corresponding influence weight-influence factor table is obtained. Each behavior mode corresponds to a different influence weight-influence factor table. For example, for the same heart rate value, the corresponding weight is different when the user is in a quiet sleep state and in a motion state. Subsequently, the behavior data is substituted into the influence weight-influence factor table to obtain the corresponding influence weight and influence factor. Then, the most recent physical measurement data is adjusted according to the influence factor and influence weight to obtain the predicted physical data, thereby achieving fine-tuning of the most recent physical measurement data according to the user's behavior. The present invention can estimate the physical condition of the current user under the state corresponding to the behavior data, and then estimate the predicted physical data, determine whether there will be health risks, and achieve estimation of the patient's physical condition after discharge, as well as early prediction of the physical condition. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 This is a flow chart of a method for predicting the risk of coronary heart disease provided by the first embodiment of the present invention;

[0048] Figure 2 This is a module diagram of a system for predicting the risk of coronary heart disease provided by the second embodiment of the present invention;

[0049] Figure 3 yes Figure 1 Schematic diagram of the process of step 101 in FIG. DETAILED DESCRIPTION

[0050] The present application will be further described in detail below in conjunction with the accompanying drawings. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application.

[0051] In the following description, for the purpose of explanation, many specific details are set forth in order to provide a thorough understanding of the inventive concepts. Some of the figures in the drawings of the present disclosure, which are part of this specification, represent structures and devices in block diagram form to avoid making the disclosed principles complicated and obscure. For the sake of clarity, not all features of an actual implementation are necessarily described. References in this disclosure to "an implementation" or "an implementation" mean that the specific features, structures or characteristics described in conjunction with that implementation are included in at least one implementation, and multiple references to "an implementation" or "an implementation" should not be understood to necessarily all refer to the same implementation.

[0052] Unless expressly limited, the terms "a", "an" and "the" are not intended to refer to a singular entity, but rather to include a general class of which a specific example may be used for illustration. Thus, the use of the term "a" or "an" may mean any number of at least one, including "one", "one or more", "at least one", and "one or more than one". The term "or" means any of the alternatives and any combination of the alternatives, including all, unless the alternatives are expressly indicated to be mutually exclusive. The phrase "at least one of" when combined with a list of items refers to a single item in the list or any combination of the items in the list. The phrase does not require all of the listed items unless expressly limited to that.

[0053] First embodiment:

[0054] The first embodiment of the present invention provides a method for predicting the risk of coronary heart disease, which obtains the user's behavioral data after physical examination and determines the user's behavior pattern based on the behavioral data; obtains the corresponding influence weight and influence factor based on the real-time detection data under the user's behavior pattern; adjusts the most recent physical examination data based on the influence factor and influence weight to obtain predicted physical data; and activates an alarm when the predicted physical data meets the preset health risk conditions.

[0055] Based on the user's behavioral data, the current user's behavioral pattern is determined, and the corresponding influence weight-influence factor table is obtained. Each behavioral pattern corresponds to a different influence weight-influence factor table. For example, for the same heart rate value, the corresponding weight is different when the user is in a quiet sleep state and in a motion state. Subsequently, the behavioral data is substituted into the influence weight-influence factor table to obtain the corresponding influence weight and influence factor. Then, the most recent physical measurement data is adjusted according to the influence factor and influence weight to obtain the predicted physical data, so as to achieve fine-tuning of the most recent physical measurement data according to the user's behavior, thereby estimating the physical condition in the state corresponding to the current user's behavioral data, and then estimating the predicted physical data, judging whether there will be health risks, and estimating the patient's physical condition after discharge, as well as making early predictions about the physical condition.

[0056] The following is a detailed description of the implementation details of a method for predicting the risk of coronary heart disease in this embodiment. The following content is only for the convenience of understanding the implementation details, and is not necessary for the implementation of this solution. The specific process of this embodiment is as follows: Figure 1 As shown, the following steps are included:

[0057] Step 101: Obtain the user's behavioral data after the physical test.

[0058] Specifically, a physical exam is a measurement of a user's physical condition or other treatment-related data performed by a hospital. In this case, the device reads the user's provided physical exam data and the time of the exam. It also collects the user's behavioral data, including location, pulse, arterial oxygen level, and number of steps. After determining the physical exam data and time, the device collects behavioral data after the exam.

[0059] In some examples, such as Figure 3 As shown, step 101 includes:

[0060] S1-1, periodic collection of behavioral data and behavior time;

[0061] S1-2, extract all behavioral time and behavioral data after the most recent physical measurement time to form the first sequence set;

[0062] S1-3, extracting the behavior time and behavior data whose difference from the current time in the first sequence set is within the first time range to form a second sequence set;

[0063] S1-4, when the second sequence set is coherent, the behavioral data in the second sequence set is used as the behavioral data after the physical test;

[0064] S1-5, when the second sequence set is incoherent, calculate the ratio of the blank behavior data in the second sequence set to the total data in the second sequence set to obtain the missing ratio;

[0065] S1-6, when the missing ratio exceeds a preset effective range, generating a first monitoring alarm signal; when the missing ratio is within the preset effective range, identifying blank behavior data based on the behavior data in the second sequence set;

[0066] S1-7, when the second sequence set is incoherent, reading the behavior time and behavior data within the second time range before and after the behavior time corresponding to the blank behavior data, and forming the third sequence set and the fourth sequence set respectively;

[0067] S1-8, predicting theoretical data of the blank behavior data according to the trends of the third sequence set and the fourth sequence set, and using the theoretical data to update the blank behavior data, thereby completing the second sequence set and making it coherent.

[0068] In S1-1, the collected behavioral data and behavior time are set by the device used in this method. This can be periodic collection after the device is started, according to a manufacturer-preset cycle; periodic collection can be periodic collection after the device is started, according to a user-defined cycle; or periodic acquisition can be performed after the measurement time and data are read. Step S1-2 is intended to form a first sequence set, which is used to filter all obtained behavioral data and behavior time. In step S1-3, data within the first time range before the current time is directly extracted and populated into a second sequence set. This step is intended to filter time segments. A determination is made as to whether all behavioral times in the second sequence set are continuous. If so, S1-4 is executed. If not, indicating the presence of blank behavioral data, S1-5 is executed. In S1-4, all behavioral data in the second sequence set is directly used as the overall behavioral data after step 101, and subsequent steps such as step 102 are performed. In S1-5, if the proportion of blank behavioral data in the second sequence set is determined, if it is relatively high, S1-6 is executed, i.e., an immediate alarm is issued; if not, S1-7 is executed. In S1-7, the data in a period of time before the blank behavior data is directly set as the third sequence set, and the data in a period of time after the blank behavior data is set as the fourth sequence set. Then, in S1-8, the blank behavior data is completed and updated according to the direction of the third sequence set and the fourth sequence set.

[0069] In a further example, in S1-8, theoretical data of blank behavior data is predicted based on the trends of the third sequence set and the fourth sequence set, including:

[0070] S1-8-1, fitting the data of the third and fourth sequence sets to obtain the initial trend model;

[0071] S1-8-2: Remove the discrete data in the third and fourth sequence sets that are at a warning distance from the initial strike, update the third and fourth sequence sets, and re-estimate the intermediate strike model based on the updated third and fourth sequence sets;

[0072] S1-8-3, fitting the initial trend model and the intermediate trend model to obtain the final trend model;

[0073] S1-8-4, calculate the theoretical data corresponding to the blank behavior data based on the final trend model.

[0074] In this example, the data in the third and fourth sequence sets are screened according to the initial trend model, and an intermediate trend model is obtained, which is finally fitted to achieve a summary of the trend patterns of the third and fourth sequence sets.

[0075] In addition, the trends of the third and fourth sequence sets represented in S1-8 in this case can also be filtered out by clustering algorithm rules, and then the fitting operation is performed.

[0076] Step 102: Determine the user behavior pattern based on the behavior data.

[0077] Specifically, step 102 includes:

[0078] S2-1, when the behavior data meets the sleep pattern, the user behavior mode is set to sleep;

[0079] S2-2, when the behavior data meets the rules of vigorous exercise, the user behavior mode is set to vigorous exercise;

[0080] S2-3, when the behavior pattern meets the regular exercise rules, the user behavior pattern is set to regular exercise.

[0081] In this example, sleep patterns, vigorous exercise patterns, and regular exercise patterns respectively illustrate the behavioral data patterns of these three user behavior modes.

[0082] Step 103: derive corresponding influence weights and influence factors based on the real-time detection data under the user behavior pattern.

[0083] Specifically, step 103 includes:

[0084] S3-1, the corresponding data weight rule is obtained according to the user behavior pattern;

[0085] S3-2, according to each influencing factor and data weight rule in the real-time detection data, the corresponding influence weight is obtained.

[0086] In this example, each user behavior pattern corresponds to a data weighting rule, where the data weighting rules are pre-set by the system. For example, sleep mode corresponds to the first data weighting rule, intense exercise mode corresponds to the second data weighting rule, and regular exercise mode corresponds to the third data weighting rule. The data weighting rules reflect the impact of various behavioral data in different modes on the user's physical health.

[0087] Then, the real-time monitoring data of the currently collected user is substituted into the data weight rule corresponding to the behavior pattern obtained based on the user's recent history, so as to obtain the influence weight corresponding to each influencing factor.

[0088] Step 104 : adjusting the most recent measured physical data according to the influencing factors and the influencing weights to obtain predicted physical data.

[0089] Specifically, the most recent physical measurement data usually corresponds to the most recent physical measurement time, and the physical measurement data is usually input by the user, and the system re-reads it from the connected hospital system, and so on.

[0090] In some examples, step 104 includes:

[0091] S4-1, substituting the impact weight into the initial impact model corresponding to the impact factor to obtain the current impact model;

[0092] S4-2, using the existing influence model to adjust the physical data of the physical test to obtain the individual physical data corresponding to the influence factor;

[0093] S4-3, based on the individual physical data corresponding to all influencing factors, the predicted physical data are comprehensively obtained.

[0094] In this example, each influencing factor corresponds to an initial influencing model, all of which are pre-installed in the system memory. Based on the influencing factor, the weight is substituted into the corresponding initial influencing model to obtain the existing influencing model. The measured body data is then adjusted based on this model to obtain the individual body data corresponding to the influencing factor. Finally, all individual body data are combined to obtain the predicted body data. The existing influencing model represents the pattern by which the weight of the influencing factor affects the measured body data. In step S4-2, a single prediction is directly performed to obtain the individual body data.

[0095] In other examples, S4-2 is replaced by integrating existing impact models corresponding to all impact factors to obtain a predicted impact model; S4-3 is replaced by using the predicted impact model to adjust the physical measurement data to obtain predicted physical data.

[0096] In this example, the influence weights of the influencing factors on the changes in the physical test data are first synthesized to obtain a prediction influence model for the overall change in the physical data. Then, the individual physical data are changed according to the prediction influence model to obtain the predicted physical data.

[0097] Step 105: When the predicted physical data meets the preset health risk condition, an alarm is activated.

[0098] Specifically, step 105 includes: if any item of the predicted physical data is within a preset risk range, then the predicted physical data meets the health risk.

[0099] In this example, physical data is predicted to determine health risks. Each health risk corresponds to a preset risk range. When each data item is within the preset risk range, it is determined that the influencing factor corresponding to the preset risk range is abnormal and an alarm is issued.

[0100] In some examples, the health risk condition includes a plurality of preset risk ranges.

[0101] Then, based on all the influencing factors that meet the preset risk range, it determines which health risk conditions may exist. For example, factors such as a rapid pulse and a weak pulse are used to determine the physical signs that meet the health risk conditions. Then, an alarm is issued. Different health risk conditions correspond to different alarm signals, and the user's corresponding guardian is notified by sending a message, etc., to remind the guardian to take the elderly person to the hospital for a check-up in time. Health risk conditions are usually determined by the diagnostic signs of the disease that needs to be monitored. For example, when coronary heart disease monitoring is usually performed, the health risk condition is determined to be the diagnostic data condition of coronary heart disease.

[0102] This method can be used to estimate the patient's physical condition after discharge, predict their condition in advance, and notify the monitoring staff to seek medical attention in a timely manner. The software corresponding to this method can be loaded into the servers of multiple home monitoring devices to enable vital sign monitoring outside the hospital.

[0103] The steps of the above-mentioned methods are divided only for the purpose of clear description. When implemented, they can be combined into one step or some steps can be split and decomposed into multiple steps. As long as they include the same logical relationship, they are all within the scope of protection of this patent. Adding insignificant modifications or introducing insignificant designs to the algorithm or process without changing the core design of the algorithm and process are all within the scope of protection of this patent.

[0104] Second embodiment:

[0105] The second embodiment of the present invention provides a system 200 for predicting the risk of coronary heart disease. Figure 2 As shown, the system includes:

[0106] The information collection module 201 is used to obtain the user's behavior data and behavior time after the physical test;

[0107] A behavior pattern determination module 202 is configured to determine a user behavior pattern based on the behavior data;

[0108] The body data deduction module 203 is used to obtain corresponding influence weights and influence factors based on the real-time detection data under the user behavior pattern, and adjust the most recent physical measurement data according to the influence factors and influence weights to obtain predicted physical data;

[0109] The health judgment module 204 is used to determine whether a preset health risk condition is met based on the predicted physical data, and to initiate an alarm when the predicted physical data meets the preset health risk condition.

[0110] Based on the user's behavioral data, the current user's behavioral pattern is determined, and the corresponding influence weight-influence factor table is obtained. Each behavioral pattern corresponds to a different influence weight-influence factor table. For example, for the same heart rate value, the corresponding weight is different when the user is in a quiet sleep state and in a motion state. Subsequently, the behavioral data is substituted into the influence weight-influence factor table to obtain the corresponding influence weight and influence factor. Then, the most recent physical measurement data is adjusted according to the influence factor and influence weight to obtain the predicted physical data, so as to achieve fine-tuning of the most recent physical measurement data according to the user's behavior, thereby estimating the physical condition in the state corresponding to the current user's behavioral data, and then estimating the predicted physical data, judging whether there will be health risks, and estimating the patient's physical condition after discharge, as well as making early predictions about the physical condition.

[0111] The information collection module in the coronary heart disease risk prediction system 200 of this solution can be configured as a port for data communication with multiple home monitoring devices, allowing for remote data collection from discharged patients via wired or wireless communication methods. The behavior pattern determination module 202, the physical data deduction module 203, and the health determination module 204 can be configured as software devices loaded onto the server of the home monitoring device, or as hardware devices.

[0112] Other implementation details and working methods of the system for predicting the risk of coronary heart disease disclosed in this application are the same as or similar to the method for predicting the risk of coronary heart disease described above and will not be repeated here.

[0113] The above are all preferred embodiments of the present application, and are not intended to limit the scope of protection of the present application. Therefore, any equivalent changes made based on the structure, shape, and principle of the present application should be included in the scope of protection of the present application.

Claims

1. A method for predicting the risk of coronary heart disease, characterized in that: The method is applied to a coronary heart disease prediction system, and the method comprises: Obtaining user behavior data after the physical test, and determining the user's behavior pattern based on the behavior data; Based on the real-time detection data under the user behavior pattern, the corresponding influence weight and influence factor are obtained; Adjust the most recent physical measurement data according to the influencing factors and the influencing weights to obtain predicted physical data; When the predicted physical data meets a preset health risk condition, triggering an alarm; Obtain user behavior data after the physical test, including: Periodically collect behavioral data and behavior time; Extract all behavioral time and behavioral data after the most recent physical test time to form a first sequence set; Extracting behavior time and behavior data whose difference from the current time in the first sequence set is within the first time range to form a second sequence set; When the second sequence set is coherent, using the behavioral data in the second sequence set as the behavioral data after the physical test; When the second sequence set is incoherent, the behavior time and behavior data within a second time range before and after the behavior time corresponding to the blank behavior data are read to form a third sequence set and a fourth sequence set respectively; predicting theoretical data of the blank behavior data according to the trends of the third sequence set and the fourth sequence set, and using the theoretical data to update the blank behavior data, thereby completing the second sequence set and making it coherent; The theoretical data of the blank behavior data is predicted based on the trends of the third sequence set and the fourth sequence set, including: fitting the data of the second sequence set and the fourth sequence set to obtain an initial trend model; Screening out discrete data in the third and fourth sequence sets whose distance from the initial strike reaches a warning value, updating the second and fourth sequence sets, and re-estimating an intermediate strike model based on the updated third and fourth sequence sets; Fitting the initial trend model and the intermediate trend model to obtain a final trend model; Theoretical data corresponding to the blank behavior data is calculated according to the final trend model.

2. The method for predicting the risk of coronary heart disease according to claim 1, wherein: Obtaining user behavior data after the physical test, including: When the second sequence set is incoherent, calculating the ratio of blank behavior data in the second sequence set to all data in the second sequence set to obtain a missing ratio; When the missing ratio exceeds a preset effective range, generating a first monitoring alarm signal; When the missing ratio is within a preset effective range, blank behavior data is identified according to the behavior data in the second sequence set.

3. The method for predicting the risk of coronary heart disease according to claim 1, wherein: Determining a user behavior pattern based on the behavior data includes: When the behavior data satisfies the sleep pattern, setting the user behavior mode to sleep; When the behavior data meets the rules of vigorous exercise, the user behavior mode is set to vigorous exercise; When the behavior pattern meets the regular exercise rules, the user behavior pattern is set as regular exercise.

4. The method for predicting the risk of coronary heart disease according to claim 1, wherein: Based on the real-time detection data under the user behavior pattern, the corresponding influence weight and influence factor are obtained, including: The corresponding data weight rules are obtained according to user behavior patterns; According to each influencing factor in the real-time detection data and the data weight rule, the corresponding influence weight is obtained.

5. The method for predicting the risk of coronary heart disease according to claim 1, wherein: According to the influencing factors and influencing weights, the most recent physical test data is adjusted to obtain the predicted physical data including: Substituting the impact weight into the initial impact model corresponding to the impact factor to obtain an existing impact model; Using the existing influence model to adjust the physical data of the physical test to obtain individual physical data corresponding to the influence factor; The predicted physical data is obtained comprehensively based on the individual physical data corresponding to all influencing factors.

6. The method for predicting the risk of coronary heart disease according to claim 1, wherein: When the predicted physical data meets a preset health risk condition, an alarm is triggered, including: If any one of the predicted physical data is within a preset risk range, the predicted physical data satisfies the health risk.

7. A system for predicting the risk of coronary heart disease, characterized in that: The system comprises: Information collection module, used to obtain the user's behavior data and behavior time after the physical test; A behavior pattern determination module, configured to determine a user's behavior pattern based on the behavior data; A body data deduction module is used to derive corresponding influence weights and influence factors based on real-time detection data under user behavior patterns, and adjust the most recent measured body data based on the influence factors and influence weights to obtain predicted body data; A health judgment module, configured to determine whether a preset health risk condition is met based on the predicted physical data, and to initiate an alarm when the predicted physical data meets the preset health risk condition; The user's behavioral data after the physical test is obtained, including: Periodically collect behavioral data and behavior time; Extract all behavioral time and behavioral data after the most recent physical test time to form a first sequence set; Extracting behavior time and behavior data whose difference from the current time in the first sequence set is within the first time range to form a second sequence set; When the second sequence set is coherent, using the behavioral data in the second sequence set as the behavioral data after the physical test; When the second sequence set is incoherent, the behavior time and behavior data within a second time range before and after the behavior time corresponding to the blank behavior data are read to form a third sequence set and a fourth sequence set respectively; predicting theoretical data of the blank behavior data according to the trends of the third sequence set and the fourth sequence set, and using the theoretical data to update the blank behavior data, thereby completing the second sequence set and making it coherent; The theoretical data of the blank behavior data is predicted based on the trends of the third sequence set and the fourth sequence set, including: fitting the data of the second sequence set and the fourth sequence set to obtain an initial trend model; Screening out discrete data in the third and fourth sequence sets whose distance from the initial strike reaches a warning value, updating the second and fourth sequence sets, and re-estimating an intermediate strike model based on the updated third and fourth sequence sets; Fitting the initial trend model and the intermediate trend model to obtain a final trend model; Theoretical data corresponding to the blank behavior data is calculated according to the final trend model.

Images