Health index monitoring method and system

Abstract

The embodiment of the invention provides a health index monitoring method and system, and relates to the technical field of computers. The health index monitoring method comprises the steps of obtaining a detection result of a health index of a user; for each health index of the user, based on user personal attribute information corresponding to the health index, a detection result of the health index and a baseline adjustment model corresponding to the health index, obtaining a current index expected range of the health index; and determining whether the health index is abnormal or not based on the current detection result of the health index and the current index expected range corresponding to the health index. According to the invention, the personalized dynamic index expected range is used for the health indexes of different users, so that the abnormality judgment of the health indexes can reflect the current health condition of an individual, the judgment accuracy of the abnormal indexes is improved, misjudgment caused by a fixed index expected range is avoided, and the user experience is improved. And the health index of the user can reflect the individual health condition more accurately.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a health indicator monitoring method and system. Background Art

[0002] In today's medical field, health management and disease prevention are receiving more and more attention. Traditional health management mainly relies on individuals to undergo periodic physical examinations for regular diagnosis; however, this method is costly and inefficient, and it is impossible to fully understand the impact of individual lifestyle and environmental factors on health in daily life.

[0003] With the development of sensor technology and big data analysis, smart healthcare is gradually coming to the fore, especially in the field of home healthcare, which allows users to perform some routine checks and health analyses at home. For example, home-use blood glucose meters, blood pressure meters, urine testers, etc. can detect common indicators such as blood sugar, blood pressure, and urine, which can reflect the user's physical condition.

[0004] However, these routine examinations and health analyses are only based on the health status of the general population, while ignoring the differences between individual family members. They cannot adapt to fluctuations in individual health status that occur over time and with changes in lifestyle, and are prone to misjudgment. Summary of the invention

[0005] The purpose of the present invention is to provide a health indicator monitoring method and system, which uses personalized dynamic indicator expected ranges for health indicators of different users, so that abnormal judgments of health indicators can better reflect the current health status of the individual, improve the accuracy of judgments of abnormal indicators, avoid misjudgments caused by fixed indicator expected ranges, and enable users' health indicators to more accurately reflect individual health status.

[0006] To achieve the above-mentioned objectives, the present invention also provides a health indicator monitoring method, including: obtaining the test results of the user's health indicators; for each of the user's health indicators, based on the user's personal attribute information corresponding to the health indicator, the test results of the health indicator and the baseline adjustment model corresponding to the health indicator, obtaining the current indicator expected range of the health indicator; based on the current test results of the health indicator and the current indicator expected range corresponding to the health indicator, determining whether the health indicator is abnormal.

[0007] The present invention also provides a health index monitoring system, comprising: a data acquisition device and a computer device in communication connection; the data acquisition device is used to obtain the detection results of the user's health index; the computer device is used to execute the above-mentioned health index monitoring method.

[0008] In one embodiment, the health index includes: a first health index that is linearly correlated with the user's personal attribute information;

[0009] The method of obtaining, for each health indicator of the user, a current expected indicator range of the health indicator based on the user personal attribute information corresponding to the health indicator, the detection result of the health indicator, and the baseline adjustment model corresponding to the health indicator, includes:

[0010] For the first health indicator, inputting the user personal attribute information corresponding to the first health indicator into a baseline adjustment model corresponding to the first health indicator to obtain a predicted value of the first health indicator;

[0011] Based on the historical detection results of the first health indicator, obtaining a fluctuation range value of the first health indicator;

[0012] Based on the predicted value of the first health indicator and the fluctuation range value of the first health indicator, the current indicator expected range of the first health indicator is obtained.

[0013] In one embodiment, for a first health indicator among the health indicators that is linearly correlated with the user's personal attribute information, the baseline adjustment model corresponding to the first health indicator is a multivariate linear regression model; the baseline adjustment model corresponding to the first health indicator is obtained in the following manner:

[0014] Acquire a first training data set related to the first health indicator;

[0015] Constructing a multiple linear regression model with the user's personal attribute information corresponding to the first health indicator as input and the first health indicator as output;

[0016] The multivariate linear regression model is trained using the first training data set to obtain a baseline adjustment model corresponding to the first health indicator.

[0017] In one embodiment, the health indicator includes: a second health indicator that has no linear correlation with the user's personal attribute information;

[0018] The method of obtaining, for each health indicator of the user, a current expected indicator range of the health indicator based on the user personal attribute information corresponding to the health indicator, the detection result of the health indicator, and the baseline adjustment model corresponding to the health indicator, includes:

[0019] For the second health indicator, the user personal attribute information corresponding to the second health indicator and the current detection result of the second health indicator are input into the baseline adjustment model corresponding to the second health indicator to obtain the current indicator expected range of the second health indicator.

[0020] In one embodiment, for a second health indicator in the health indicators that has no linear correlation with the user's personal attribute information, the baseline adjustment model corresponding to the second health indicator is a random forest regression model; the baseline adjustment model corresponding to the second health indicator is obtained by:

[0021] Acquire a second training data set related to a second health indicator;

[0022] Constructing a random forest regression model, wherein the random forest regression model includes a plurality of decision trees, each of which takes a user's personal attribute corresponding to the second health indicator and a historical detection result of the second health indicator as input and takes a prediction indicator range of the second health indicator as output;

[0023] Using the second training data set to train each of the decision trees in the random forest regression model;

[0024] The multiple trained decision trees are combined to obtain a baseline adjustment model corresponding to the second health indicator.

[0025] In one embodiment, the health indicator includes: a third health indicator related to historical test results in a recent set time period;

[0026] The method of obtaining, for each health indicator of the user, a current expected indicator range of the health indicator based on the user personal attribute information corresponding to the health indicator, the detection result of the health indicator, and the baseline adjustment model corresponding to the health indicator, includes:

[0027] For the third health indicator, the user's personal attribute information corresponding to the third health indicator and the historical detection result of the third health indicator in a set time period are input into the baseline adjustment model corresponding to the three health indicators to obtain the predicted value of the third health indicator in the future specified time period;

[0028] Determining a baseline adjustment parameter of the third health indicator based on a predicted value of the third health indicator within a specified future time period;

[0029] The indicator range of the third health indicator is adjusted based on the baseline adjustment parameter to obtain a current indicator expected range of the third health indicator.

[0030] In one embodiment, for the third health indicator related to the historical detection results of the most recent set time period in the health indicators, the baseline adjustment model corresponding to the third health indicator is a time series analysis model; the baseline adjustment model corresponding to the third health indicator is obtained in the following manner:

[0031] Obtaining a third training data set corresponding to the third health indicator;

[0032] Constructing a time series analysis model that takes the historical detection results of the third training data within a set time period and the corresponding time points as input and takes the predicted value of the third health indicator within a future specified time period as output, wherein the time series analysis model is an autoregressive integrated moving average model or a long short-term memory neural network model;

[0033] The time series analysis model is trained using the third training data set to obtain a baseline adjustment model corresponding to the third health indicator.

[0034] In one embodiment, the health indicators include: urine test indicators;

[0035] The urine detection index includes at least one of the following: uric acid, creatinine, albumin, and ketone bodies.

[0036] In one embodiment, the health indicators include: blood test indicators;

[0037] The blood test index includes at least one of the following: blood pressure, blood lipids, and blood sugar.

[0038] In one embodiment, the user personal attribute information includes at least one of the following: height, gender, age, weight, living habit information, eating habit information, and living environment information. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 is a specific flow chart of the health indicator monitoring method according to the first embodiment of the present invention. DETAILED DESCRIPTION

[0040] The following will be described in detail with reference to the accompanying drawings to provide a clearer understanding of the purpose, features and advantages of the present invention. It should be understood that the embodiments shown in the accompanying drawings are not intended to limit the scope of the present invention, but are only intended to illustrate the essential spirit of the technical solution of the present invention.

[0041] In the following description, certain specific details are set forth for the purpose of illustrating various disclosed embodiments to provide a thorough understanding of the various disclosed embodiments. However, those skilled in the relevant art will recognize that the embodiments may be practiced without one or more of these specific details. In other cases, well-known devices, structures, and techniques associated with the present application may not be shown or described in detail to avoid unnecessarily obscuring the description of the embodiments.

[0042] Unless the context requires otherwise, throughout the specification and claims, the word "comprise" and variations such as "include" and "have" should be construed in an open, inclusive sense, ie, should be interpreted as "including, but not limited to."

[0043] References throughout the specification to "one embodiment" or "an embodiment" indicate that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment. Thus, the appearances of "in one embodiment" or "in an embodiment" in various places throughout the specification are not necessarily all referring to the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any manner in one or more embodiments.

[0044] As used in this specification and the appended claims, the singular forms "a", "an", and "the" include plural referents unless the context clearly dictates otherwise. It should be noted that the term "or" is generally employed in its sense including "or / and" unless the context clearly dictates otherwise.

[0045] In the following description, in order to clearly show the structure and working mode of the present invention, many directional words will be used for description, but the words "front", "back", "left", "right", "outside", "inside", "outward", "inward", "up", "down", etc. should be understood as convenient terms and should not be understood as restrictive terms.

[0046] The first embodiment of the present invention relates to a health index monitoring method, which is applied to computer devices such as mobile phones, desktop hosts, and laptop computers. The health index detection method is mainly used in fixed places such as families, and this embodiment is described by taking one user as an example.

[0047] The specific process of the health indicator monitoring method in this embodiment is as follows: Figure 1 shown.

[0048] Step 101, obtaining the test results of the user's health indicators.

[0049] Specifically, the test results of the user's health indicators can be actively uploaded to the computer device by the user after the test, or the computer device can be connected to the data acquisition device, and the data acquisition device can acquire relevant data. Furthermore, the data acquisition device can include some detection devices, such as a sphygmomanometer for detecting blood pressure, a blood glucose meter for detecting blood sugar, or a urine detection device for collecting the user's urine for testing. The urine detection device can be installed in the toilet, so that the user's urine can be collected for testing and the user's urine test results can be obtained. Urine testing refers to the detection of urine in urine. Routine urine test results include, but are not limited to: urine acidity (pH), urine white blood cells (WBC / LEU), urine nitrite (NIT), urine protein (PRO), glucose (GLU), urobilinogen (URO / UBG), ketone bodies (KET), occult blood (BLU), bilirubin (BIL), ascorbic acid (VC), salts, creatinine (CrE), and may also include indicators of tumor markers, such as alpha-fetoprotein, carcinoembryonic antigen, albumin, urine β2-microglobulin, etc., as well as microbial detection indicators, such as nicotine, ethanol, vitamin C, etc.

[0050] The computer device directly or indirectly obtains the detection results of these detection devices through wired or wireless means; for example, the detection device is directly connected to the computer device for communication, so that the detection device can directly send the detection results to the computer device after the user completes the detection using the detection device, and the detection device can also perform detection or send the detection results based on the user's instructions; for another example, after the user completes the detection using the detection device, the detection device uploads the data to the cloud server, and the computer device obtains the detection results from the cloud server.

[0051] The storage method of the user's test results for each health indicator is: user name ID, user personal attribute information, and the test results of each health indicator and the corresponding test time. Among them, the user's name ID and user personal attribute information can be pre-stored in the computer device in advance, and are established and stored by the user when using the computer device for the first time.

[0052] Before or after each test, the user can associate his or her own user name with the test result. For example, the detection device includes an identity recognition device, which can be an NFC sensor or a biometric sensor (fingerprint, face, iris, etc.). When the user uses it, the user's identity can be recognized, the user name is automatically associated with the test result, and the test time is recorded.

[0053] Regardless of the method used, the test results stored in the computer device are all data that corresponds to the information of user name-personal attribute information-health index-test result-test time.

[0054] The health indicators here include: common blood test indicators, such as at least one of the following: blood pressure, blood lipids, blood sugar. The blood test indicators may also include other blood test indicators that are of certain significance to individual users, such as erythrocyte sedimentation rate.

[0055] Health indicators may also include: common urine test indicators, including at least one of the following: uric acid, creatinine, albumin, ketone bodies. Urine test indicators may also include other urine test indicators that are of certain significance to individual users, such as HCG, urobilinogen, sodium nitrite, etc.

[0056] The user personal attribute information corresponding to each user includes at least one of the following: height, gender, age, weight, eating habits information, and living environment information. The eating habits information may be, for example, sweets consumption preference (divided into three levels: high, medium and low), the living habits information may include: exercise preference, work and rest time, etc., and the living environment information may include indoor temperature, humidity and air quality information, etc. The eating habits information and living habits information may be filled in and updated by the user himself, and the living environment information may be collected and updated in real time through indoor temperature and humidity sensors, air quality sensors, etc.

[0057] The above-mentioned health indicator test results, user personal attribute information and other information are pre-processed data. For example, for missing data (for example, certain health indicators are missing in the last test), the mean, interpolation, nearest neighbor interpolation and other methods can be used to complete the data; numerical data are standardized so that data of different scales can be consistent in the baseline adjustment model; for some health indicators (such as eating habits and exercise habits), key features are extracted so that these relevant information can be converted into a format suitable for baseline adjustment model input.

[0058] Step 102, for each health indicator of the user, based on the user's personal attribute information corresponding to the health indicator, the detection result of the health indicator and the baseline adjustment model corresponding to the health indicator, obtain the current indicator expected range of the health indicator.

[0059] Specifically, the computer device stores the baseline adjustment model of each health indicator, so that for each health indicator of the current user, the user's personal attribute information corresponding to the health indicator, the historical test results of the health indicator and the baseline adjustment model corresponding to the health indicator can be used to obtain the current expected indicator range of the health indicator. In other words, the current expected indicator range of each health indicator of the user can be updated according to a set time period or based on user instructions, or each time the user uploads new test data of a health indicator, the current expected indicator range of the health indicator is updated. This can achieve dynamic adjustment of the expected indicator range of each health indicator. The expected indicator range of the health indicator is combined with the user's personal attribute information, conforms to the user's current personal status, and is more targeted.

[0060] Different types of models can be used to train different health indicators to obtain a baseline adjustment model. The following divides the health indicators into three categories for explanation. Of course, you can also choose one or more categories of health indicators according to your needs and use the same model to train a baseline adjustment model.

[0061] The first category, the health indicators include: a first health indicator that is linearly correlated with the user's personal attribute information; the first health indicator is, for example, blood pressure, and the user's personal attribute information that is linearly correlated with blood pressure includes: age, weight, gender, exercise frequency, eating habits (salt intake), etc.

[0062] For the first health indicator among the health indicators that is linearly correlated with the user's personal attribute information, the baseline adjustment model corresponding to the first health indicator is a multivariate linear regression model; the baseline adjustment model corresponding to the first health indicator is obtained as follows:

[0063] Obtain a first training data set related to the first health indicator; wherein the first training data set includes: historical detection values ​​of the user for the first health indicator within a certain period of time in the past and attribute values ​​of the user's personal attribute information related to the first health indicator under each historical detection value.

[0064] Construct a multiple linear regression model with the user's personal attribute information corresponding to the first health indicator as input and the first health indicator as output; the expression of the multiple linear regression model is, for example: Y d1 =a1*x1+a2*x2+a3*x3+a4*x4+...+a n *x n , where Y d1 Indicates the detection value of the first health indicator, x1 to x n Represents n personal attributes, a1 to a n Represents the weight of each personal attribute, where n is an integer greater than or equal to 1.

[0065] The multivariate linear regression model is trained using the first training data set to obtain a baseline adjustment model corresponding to the first health indicator. Specifically, the multivariate linear regression model is trained using the first training data set to determine the weight of the influence of each personal attribute on the first health indicator. The obtained multivariate linear regression model is the baseline adjustment model corresponding to the first health indicator.

[0066] The specific method of obtaining the expected range of the current indicator corresponding to the first health indicator is:

[0067] For the first health indicator, the user personal attribute information corresponding to the first health indicator is input into the baseline adjustment model corresponding to the first health indicator to obtain the predicted value of the first health indicator; that is to say, the current attribute value of each user personal attribute corresponding to the first health indicator is input into the baseline adjustment model corresponding to the first health indicator, so as to obtain the predicted value of the first health indicator of the user in the current state.

[0068] Based on the historical test results of the first health indicator, the fluctuation range value of the first health indicator is obtained; the historical test results at this time may refer to the fluctuation range of the first health indicator of the target group that is the same or similar to the user's various personal attributes (referring to the user's personal attributes corresponding to the first health indicator), and the standard deviation of the first health indicator of the target group is calculated or the fluctuation range value is given in the form of quantiles.

[0069] Based on the predicted value of the first health indicator and the fluctuation range of the first health indicator, the expected range of the current indicator of the first health indicator is obtained. For example, the predicted value of the first health indicator is K1, the fluctuation range of the first health indicator is ±5%, and the predicted value K1 of the first health indicator is taken as the center value, then the expected range of the current indicator of the first health indicator is [A-5%*A, A+5%*A].

[0070] The second category, the health indicators include: a second health indicator that has no linear correlation with the user's personal attribute information. The second health indicator is, for example, a health indicator such as uric acid and blood sugar that is affected by multiple factors.

[0071] For a second health indicator among the health indicators that has no linear correlation with the user personal attribute information, the baseline adjustment model corresponding to the second health indicator is a random forest regression model; and the method for obtaining the baseline adjustment model corresponding to the second health indicator is:

[0072] Obtain a second training data set related to the second health indicator; wherein the second training data set includes: historical detection values ​​of the user for the second health indicator within a certain period of time in the past, attribute values ​​of the user's personal attribute information related to the second health indicator under each historical detection value, and baseline prediction values ​​corresponding to each historical detection value. The baseline prediction value here can be the usual range of the second health indicator for a group of similar personal attributes obtained through big data statistics, thereby presetting corresponding baseline prediction values ​​for the historical detection values ​​of each second health indicator; or a multivariate linear regression model can be first established for the second health indicator and the user's personal attribute information related to it, and then the baseline prediction values ​​of each second health indicator under each historical detection value are obtained. The specific process is similar to the above and will not be repeated here. In addition, a test set is generally obtained in advance to test the trained model.

[0073] A random forest regression model is constructed, wherein the random forest regression model includes a plurality of decision trees, each of which takes the user's personal attribute corresponding to the second health indicator and the historical detection result of the second health indicator as input, and takes the predicted indicator range of the second health indicator as output. Different decision trees are divided according to various personal attributes related to the second health indicator, and one or more features are selected as the segmentation basis of each decision tree; for example, a decision tree based on age is divided according to different age groups.

[0074] The second training data set is used to train each decision tree in the random forest regression model; the multiple decision trees that have been trained are combined to obtain a baseline adjustment model corresponding to the second health indicator. Different data in the second training data set are selected for training different decision trees to obtain each trained decision tree, and all the decision trees after training are integrated to obtain a random forest regression model. The outputs of all decision trees are voted or averaged to obtain the final result; then, cross-validation can be used to ensure the generalization ability of the random forest regression model and optimize the number and depth of decision trees in the random forest regression model; the random forest regression model can also be verified using data in the test set, and the model parameters can be adjusted according to the verification results to eliminate fitting or underfitting problems.

[0075] The specific method of obtaining the expected range of the current indicator corresponding to the second health indicator is:

[0076] For the second health indicator, the user's personal attribute information corresponding to the second health indicator and the historical detection results of the second health indicator are input into the baseline adjustment model corresponding to the second health indicator to obtain the current expected range of the second health indicator; the baseline adjustment model corresponding to the second health indicator is a random forest regression model, and its output is the current expected range of the second health indicator after dynamic adjustment; thus, the user's personal attribute information corresponding to the second health indicator and the current detection results of the second health indicator can be input into the baseline adjustment model corresponding to the second health indicator, and the baseline adjustment model will output the current expected range of the second health indicator.

[0077] The third category, the health indicators include: the third health indicators related to the historical test results of the most recent set time period; that is, the third health indicators are health indicators with time correlation, such as health indicators that change regularly due to seasons or daily life cycles.

[0078] For the third health indicator related to the historical detection results of the most recent set time period in the health indicators, the baseline adjustment model corresponding to the third health indicator is a time series analysis model; the baseline adjustment model corresponding to the third health indicator is obtained in the following manner:

[0079] A third training data set corresponding to the third health indicator is obtained; wherein the third training data set includes historical test results of the user's third health indicator and the corresponding test time points, as well as the user's personal attribute information at each test time point. The data in the third training data set is a decomposition of the time series, from which trend items, seasonal items and residual items are separated to facilitate model training processing.

[0080] A time series analysis model is constructed with the historical detection results of the third training data within a set time period and the corresponding time points as input and the predicted values ​​of the third health indicator within a specified future time period as output. The time series analysis model is an autoregressive integrated moving average model or a long short-term memory neural network model. The autoregressive integrated moving average model is suitable for linear time series, that is, the scenario of regular measurement of the third health indicator, and the long short-term memory neural network model is suitable for nonlinear and complex time series, that is, the scenario of random measurement of the third health indicator. The appropriate time series analysis model can be selected according to the measurement scenario of the third health indicator.

[0081] The constructed time series analysis model takes the historical test results and their test time in the past set time period, and the attribute values ​​of the user's personal attributes (related to the third health indicator) at each test time as input, and takes the predicted value of the third health indicator at each time point in the future specified time period as output. When the time series analysis model is trained using the training data in the third training set, for each training data, data of the continuous set time period + the length of the specified time period are selected, and then the time series analysis model is trained.

[0082] The time series analysis model is trained using the third training data set to obtain a baseline adjustment model corresponding to the third health indicator. For example, for an autoregressive integrated moving average model, the p, d, and q parameters can be obtained after training.

[0083] The specific method of obtaining the expected range of the current indicator corresponding to the third health indicator is:

[0084] For the third health indicator, the user personal attribute information corresponding to the third health indicator and the historical detection results of the third health indicator in the set time period are input into the baseline adjustment model corresponding to the three health indicators to obtain the predicted value of the third health indicator in the future specified time period; wherein the historical detection results of the set time period may be based on this detection as the starting point, and the historical inspection results of the set time period selected forward may include the current detection results of the third health indicator; the baseline adjustment model corresponding to the third health indicator outputs the predicted value of the third health indicator in the future specified time period.

[0085] Based on the predicted value of the third health indicator in the future specified time period, the baseline adjustment parameter of the third health indicator is determined; specifically, based on the predicted value of the third health indicator in the future specified time period, the change trend and change range of the third health indicator in a certain time period in the future can be determined. For example, the third health indicator gradually increases in the future specified time period. First, a linear function of the third health indicator at each time point in the future specified time period and the change range is fitted. Based on this, the change range value of the third health indicator at each time point in the future specified time period can be determined based on the linear function. For example, if the increase range of the third health indicator at a certain time point in the future specified time period is 3%, then the baseline adjustment parameter of the third health indicator at this time point is +3%. It should be noted that the change range value of the third health indicator at each time point in the future specified time period is updated each time the test is completed or according to a certain period to ensure the accuracy of the baseline adjustment parameter.

[0086] The indicator range of the third health indicator is adjusted based on the baseline adjustment parameter to obtain the expected current indicator range of the third health indicator. Since the change amplitude value of the third health indicator at each time point in the future specified time period has been obtained, after the detection of the third health indicator at a certain time point is completed, the baseline adjustment parameter of the time point can be directly obtained, and then the indicator range of the previous third health indicator is adjusted based on the baseline adjustment parameter to obtain the expected current indicator range of the third health indicator.

[0087] For example, if the indicator range of the third health indicator before adjustment is [k3, k4], and the current baseline adjustment parameter of the third health indicator is +3%, then the current expected indicator range of the third health indicator is [k3*1.03, k4*1.03]. For example: if it is predicted that the user's blood sugar will gradually rise in the future, the blood sugar baseline can be gradually increased by a certain proportion to make it consistent with the future health status; for example, if it is predicted that blood pressure is higher in winter, the blood pressure baseline can be appropriately increased during the winter.

[0088] Step 103, determining whether the health indicator is abnormal based on the current detection result of the health indicator and the expected range of the current indicator corresponding to the health indicator.

[0089] Specifically, after the user completes the detection of one or more health indicators and the computer device receives the latest current detection results of these health indicators, it can judge whether these health indicators are abnormal in combination with the current expected range of these health indicators, that is, compare the current detection result of each health indicator with the corresponding current expected range of the indicator to determine whether the user's health indicator is abnormal.

[0090] Furthermore, since the expected range of the current indicator is combined with the user's personal attribute information, it is also possible to determine where the problem is with abnormal health indicators combined with changes in the user's personal attribute information. For example, if the user's blood sugar is too high, and the user's eating habits have a preference for sweets and little exercise, the user can be advised to reduce the intake of sweets and increase the amount of exercise; or, if the user's blood pressure is too high and the user's eating habits have a preference for salty foods, the user can be advised to reduce the intake of high-salt foods.

[0091] In this embodiment, for each of the health indicators of the user, the current indicator expected range of the health indicator is obtained based on the user's personal attribute information corresponding to the health indicator, the test results of the health indicator and the baseline adjustment model corresponding to the health indicator. That is to say, the current indicator expected range of each health indicator of the user can be updated according to a set time period or based on user instructions, or each time the user uploads new test data of a health indicator, the current indicator expected range of the health indicator is updated, thereby achieving dynamic and real-time adjustment of the indicator expected range of each health indicator. The indicator expected range of the health indicator combines the user's personal attribute information, conforms to the user's current personal status, and is more targeted; subsequently, personalized dynamic indicator expected ranges are used for the health indicators of different users, so that the abnormal judgment of the health indicator can better reflect the individual's current health status, improve the judgment accuracy of the abnormal indicator, avoid the misjudgment caused by the fixed indicator expected range, and enable the user's health indicator to more accurately reflect the individual's health status.

[0092] The second embodiment of the present invention relates to a health index monitoring system, comprising: a data acquisition device and a computer device in communication connection; the computer device may be a mobile phone, a desktop host, a laptop computer, etc.

[0093] The data acquisition device is used to obtain the detection results of the user's health indicators.

[0094] The computer device is used to execute the health indicator monitoring method described in the first embodiment.

[0095] The test results of the user's health indicators can be actively uploaded to the computer device by the user, or the computer device can be connected to the data acquisition device, and the data acquisition device can acquire relevant data. Furthermore, the data acquisition device can include some detection devices, such as a sphygmomanometer for detecting blood pressure, a blood glucose meter for detecting blood sugar, or a urine detection device for collecting the user's urine for detection. The urine detection device can be installed in the toilet, so that the user's urine can be collected for detection and the user's urine detection results can be obtained; the computer device directly or indirectly obtains the detection results of these detection devices through wired or wireless means; for example, the detection device is directly connected to the computer device for communication, so that the detection device can directly send the detection results to the computer device after the user completes the detection using the detection device; for another example, after the user completes the detection using the detection device, the detection device uploads the data to the cloud server, and the computer device obtains the detection data from the cloud server.

[0096] Since the first embodiment corresponds to this embodiment, this embodiment can be implemented in conjunction with the first embodiment. The relevant technical details mentioned in the first embodiment are still valid in this embodiment, and the technical effects that can be achieved in the first embodiment can also be achieved in this embodiment. In order to reduce repetition, they are not repeated here. Accordingly, the relevant technical details mentioned in this embodiment can also be applied in the first embodiment.

[0097] Preferred embodiments of the present invention have been described above in detail, but it should be understood that aspects of the embodiments can be modified, if necessary, to employ aspects, features and concepts of the various patents, applications and publications to provide further embodiments.

[0098] These and other changes can be made to the embodiments in light of the above detailed description.In general, in the claims, the terms used should not be considered limited to the specific embodiments disclosed in the specification and the claims, but should be construed to include all possible embodiments along with the full scope of equivalents to which these claims are entitled.

Claims

1. A health index monitoring method, characterized in that: include: Obtain the test results of the user's health indicators; For each health indicator of the user, based on the user personal attribute information corresponding to the health indicator, the detection result of the health indicator and the baseline adjustment model corresponding to the health indicator, obtaining the current indicator expected range of the health indicator; Based on the current detection result of the health indicator and the expected range of the current indicator corresponding to the health indicator, it is determined whether the health indicator is abnormal.

2. The health indicator monitoring method according to claim 1, characterized in that: The health index includes: a first health index that is linearly correlated with the user's personal attribute information; The method of obtaining, for each health indicator of the user, a current expected indicator range of the health indicator based on the user personal attribute information corresponding to the health indicator, the detection result of the health indicator, and the baseline adjustment model corresponding to the health indicator, includes: For the first health indicator, inputting the user personal attribute information corresponding to the first health indicator into a baseline adjustment model corresponding to the first health indicator to obtain a predicted value of the first health indicator; Based on the historical detection results of the first health indicator, obtaining a fluctuation range value of the first health indicator; Based on the predicted value of the first health indicator and the fluctuation range value of the first health indicator, the current indicator expected range of the first health indicator is obtained.

3. The health indicator monitoring method according to claim 1 or 2, characterized in that: For the first health indicator among the health indicators that is linearly correlated with the user's personal attribute information, the baseline adjustment model corresponding to the first health indicator is a multivariate linear regression model; the baseline adjustment model corresponding to the first health indicator is obtained as follows: Acquire a first training data set related to the first health indicator; Constructing a multiple linear regression model with the user's personal attribute information corresponding to the first health indicator as input and the first health indicator as output; The multivariate linear regression model is trained using the first training data set to obtain a baseline adjustment model corresponding to the first health indicator.

4. The health index monitoring method according to claim 1, characterized in that: The health index includes: a second health index which has no linear correlation with the user's personal attribute information; The method of obtaining, for each health indicator of the user, a current expected indicator range of the health indicator based on the user personal attribute information corresponding to the health indicator, the detection result of the health indicator, and the baseline adjustment model corresponding to the health indicator, includes: For the second health indicator, the user personal attribute information corresponding to the second health indicator and the current detection result of the second health indicator are input into the baseline adjustment model corresponding to the second health indicator to obtain the current indicator expected range of the second health indicator.

5. The health index monitoring method according to claim 1 or 4, characterized in that: For a second health indicator among the health indicators that has no linear correlation with the user personal attribute information, the baseline adjustment model corresponding to the second health indicator is a random forest regression model; and the method for obtaining the baseline adjustment model corresponding to the second health indicator is: Acquire a second training data set related to a second health indicator; Constructing a random forest regression model, wherein the random forest regression model includes a plurality of decision trees, each of which takes a user's personal attribute corresponding to the second health indicator and a historical detection result of the second health indicator as input and takes a prediction indicator range of the second health indicator as output; Using the second training data set to train each of the decision trees in the random forest regression model; The multiple trained decision trees are combined to obtain a baseline adjustment model corresponding to the second health indicator.

6. The health index monitoring method according to claim 1, characterized in that: The health indicators include: a third health indicator related to the historical test results of the most recent set time period; The method of obtaining, for each health indicator of the user, a current expected indicator range of the health indicator based on the user personal attribute information corresponding to the health indicator, the detection result of the health indicator, and the baseline adjustment model corresponding to the health indicator, includes: For the third health indicator, the user's personal attribute information corresponding to the third health indicator and the historical detection result of the third health indicator in a set time period are input into the baseline adjustment model corresponding to the three health indicators to obtain the predicted value of the third health indicator in the future specified time period; Determining a baseline adjustment parameter of the third health indicator based on a predicted value of the third health indicator within a specified future time period; The indicator range of the third health indicator is adjusted based on the baseline adjustment parameter to obtain a current indicator expected range of the third health indicator.

7. The health index monitoring method according to claim 1 or 6, characterized in that: For a third health indicator among the health indicators that is related to the historical detection results of the most recent set time period of the health indicator, the baseline adjustment model corresponding to the third health indicator is a time series analysis model; the baseline adjustment model corresponding to the third health indicator is obtained in the following manner: Obtaining a third training data set corresponding to the third health indicator; Constructing a time series analysis model that takes the historical detection results of the third training data within a set time period and the corresponding time points as input and takes the predicted value of the third health indicator within a future specified time period as output, wherein the time series analysis model is an autoregressive integrated moving average model or a long short-term memory neural network model; The time series analysis model is trained using the third training data set to obtain a baseline adjustment model corresponding to the third health indicator.

8. The health index monitoring method according to claim 1, characterized in that: The health indicators include: urine test indicators; The urine detection index includes at least one of the following: uric acid, creatinine, albumin, and ketone bodies.

9. The health index monitoring method according to claim 1 or 8, characterized in that: The health indicators include: blood test indicators; The blood test index includes at least one of the following: blood pressure, blood lipids, and blood sugar.

10. The health index monitoring method according to claim 1, characterized in that: The user's personal attribute information includes at least one of the following: height, gender, age, weight, living habit information, eating habit information, and living environment information.

11. A health index monitoring system, characterized in that: include: A data acquisition device and a computer device in communication connection; The data acquisition device is used to obtain the detection results of the user's health indicators; The computer device is used to execute the health indicator monitoring method according to any one of claims 1 to 10.

12. The health indicator monitoring system according to claim 11, characterized in that: The data acquisition device includes: a urine detection device; the urine detection device is provided for assembly in a toilet; the health index includes: a urine detection index; The urine testing device is used to collect urine of a user for testing and obtain the urine testing result of the user.

Images