Early warning system and method for high-risk pregnant women

Abstract

The present invention discloses a high-risk maternal condition early warning system and method, which relates to the field of intelligent health monitoring technology. The method includes collecting individualized information and physiological data of pregnant women, automatically identifying the pregnancy stage, and generating a maternal data packet; constructing a multi-stage hierarchical risk prediction model based on historical data, and predicting the maternal stage health risk score according to the maternal data packet; combining the maternal stage health risk score with the individualized information of the pregnant women, optimizing the maternal stage health risk score through fuzzy logic, and generating a personalized maternal health risk score; evaluating the current health risk level of the pregnant women according to the personalized maternal health risk score; triggering the early warning mechanism and providing personalized intervention suggestions according to the current health risk level of the pregnant women. The present invention improves the accuracy of the early warning of the high-risk maternal condition by constructing a multi-stage hierarchical risk prediction model, and ensures the health and safety of mothers and infants.

Description

Technical Field

[0001] The present invention relates to the field of intelligent health monitoring technology, and in particular to a system and method for early warning of high-risk maternal and child illness. Background Art

[0002] With the development of modern medical technology, the means of pregnancy monitoring for pregnant women have gradually developed in the direction of digitalization and intelligence. Traditional maternal health monitoring mainly relies on regular prenatal examinations and the subjective judgment of medical staff. Although it is effective, it often has problems of insufficient timeliness and continuity. In recent years, with the advancement of technologies such as big data analysis and machine learning, more and more studies are committed to building data-driven health risk prediction models in order to detect the potential disease risks of high-risk pregnant women earlier, thereby improving the efficiency and effectiveness of intervention. In this context, many health monitoring systems based on electronic health records (EHR) and other data sources have gradually been applied in clinical practice. However, most of the existing systems and methods focus on the monitoring of a single physiological parameter, such as blood pressure or blood sugar, and lack comprehensive analysis of multi-dimensional physiological data, especially in the accurate prediction of individualized health risks of pregnant women at different stages of pregnancy. There are still many imperfections.

[0003] There are two main deficiencies in the existing related technologies. First, most high-risk maternal health monitoring systems only perform simple risk assessments based on a single or a few physiological data, and fail to fully integrate the individualized information of pregnant women (such as age, medical history, family genetic history, etc.) and multi-dimensional physiological data (such as fetal heart rate, blood pressure, blood sugar, etc.), resulting in incomplete and inaccurate assessment of maternal health risks. Secondly, in the process of risk assessment, the existing systems often ignore the time dependence of the impact of dynamic changes in the pregnancy stage on physiological data, that is, they fail to fully consider the impact of different characteristics of the pregnancy stage on maternal health risks, making it difficult to achieve effective classification and stratified assessment of different pregnancy stages. Especially in terms of early warning of the condition of high-risk pregnant women, the real-time and accuracy of the existing systems are relatively insufficient, and they cannot provide sufficient personalized intervention recommendations. Summary of the invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a system and method for early warning of high-risk maternal conditions to solve the problem in the prior art that it is difficult to dynamically assess health risks at different stages of pregnancy.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In the first aspect, an embodiment of the present invention provides a method for early warning of high-risk maternal conditions, which includes collecting individualized information and physiological data of pregnant women, automatically identifying the stage of pregnancy, and generating a data packet for pregnant women; constructing a multi-stage hierarchical risk prediction model based on historical data, and predicting the maternal stage health risk score according to the maternal data packet; combining the maternal stage health risk score with the individualized information of the pregnant women, optimizing the maternal stage health risk score through fuzzy logic, and generating a personalized maternal health risk score; evaluating the current health risk level of the pregnant women according to the personalized maternal health risk score; triggering an early warning mechanism and providing personalized intervention suggestions according to the current health risk level of the pregnant women.

[0008] As a preferred embodiment of the high-risk maternal condition early warning method of the present invention, the individualized information of the maternal condition includes age, medical history, family genetic disease history and previous pregnancy status;

[0009] The physiological data include fetal heart rate, blood pressure, blood sugar, body temperature and uterine contraction intensity.

[0010] As a preferred solution of the high-risk maternal condition early warning method of the present invention, the automatic identification of the pregnancy stage and the generation of the maternal data packet are specifically performed as follows:

[0011] Based on the last menstrual period date of the pregnant woman, the expected date of delivery is calculated using the Naegele formula, and the current stage of pregnancy of the pregnant woman is identified based on the expected date of delivery and the current date;

[0012] Integrate the collected maternal physiological data and current pregnancy stage information into a complete data package.

[0013] As a preferred solution of the high-risk maternal condition early warning method of the present invention, wherein: based on historical data, a multi-stage hierarchical risk prediction model is constructed, and the maternal stage health risk score is predicted according to the maternal data packet, and the specific steps are as follows:

[0014] Historical data were collected from the hospital's electronic health record and maternal management system database;

[0015] Label the historical data to mark the different conditions of pregnant women and the corresponding pregnancy stages;

[0016] According to the different stages of pregnancy corresponding to different conditions of pregnant women, independent assessment layers are designed for early, mid-term, late and labor periods;

[0017] The early evaluation layer uses the conditional dependency inference mechanism of the Bayesian network to capture the complex associations between physiological data in the early stages;

[0018] The mid-term evaluation layer uses the memory capacity of LSTM to capture the dynamic changes of physiological data over time in the mid-term stage;

[0019] The late-stage evaluation layer uses XGBoost’s feature selection and ensemble learning capabilities to screen out key risk factors from a large amount of physiological data in the late stage;

[0020] The labor assessment layer uses the reinforcement learning capabilities of Deep Q-Learning to capture the real-time dynamic changes of physiological data during the labor stage;

[0021] Each independent assessment layer is independently trained with physiological data from different pregnancy stages in historical data to strengthen the perception of each independent assessment layer on the physiological data of pregnant women at the corresponding stage, and ultimately forms a multi-stage hierarchical risk prediction model that can perform targeted risk prediction based on the physiological data of pregnant women at different pregnancy stages;

[0022] The maternal data package is input into the multi-stage hierarchical risk assessment model. According to the current expected delivery stage of the pregnant woman, the assessment layer is dynamically selected for risk score prediction. The expression is: ;in, Indicates the health risk score of the pregnant woman in the current pregnancy stage. represents the physiological data of pregnant women at time t, represents the hyperbolic tangent function, represents the influence of the ith physiological data on the risk score, represents the mean of the ith physiological data, represents the standard deviation of the ith physiological data, and n is the total number of physiological data.

[0023] As a preferred solution of the high-risk maternal condition early warning method of the present invention, wherein: the maternal stage health risk score is combined with the individualized information of the maternal, and the maternal stage health risk score is optimized by fuzzy logic to generate a personalized maternal health risk score. The specific steps are as follows:

[0024] Through fuzzy processing, the individual information of pregnant women and the health risk score of pregnant women in the current pregnancy stage are selected. As input variables, fuzzy sets are divided for each input variable and membership functions are defined. The precise values ​​of the input variables are mapped to the fuzzy sets through membership functions to generate fuzzy information.

[0025] Based on clinical experience and historical data, select a health risk score that is relevant to the current stage of pregnancy. Relevant individualized information of pregnant women, by defining individualized information of pregnant women and health risk scores of pregnant women in the current stage of pregnancy The causal relationship is established by setting fuzzy rules;

[0026] Match the fuzzified information with the fuzzy rules, calculate the activation degree of each rule through the Mamdani algorithm and synthesize the fuzzy output;

[0027] The synthesized fuzzy output is converted into a numerical health risk score through defuzzification, and then the health risk score of the pregnant woman in the current pregnancy stage is scored. Dynamically adjust to generate personalized maternal health risk scores .

[0028] As a preferred solution of the high-risk maternal condition early warning method of the present invention, wherein: the current health risk level of the maternal health is assessed according to the personalized maternal health risk score, and the specific steps are as follows:

[0029] Based on historical data, a low score threshold T1 and a high score threshold T2 are defined to assess the current health risk level of pregnant women.

[0030] when When ≤T1, the current maternal health risk is considered to be low;

[0031] When T1< When ≤T2, the current maternal health is considered to be at medium risk;

[0032] when >T2, the current maternal health is considered to be at high risk.

[0033] As a preferred solution of the high-risk maternal condition early warning method of the present invention, wherein: according to the current health risk level of the pregnant woman, the early warning mechanism is triggered and personalized intervention suggestions are provided. The specific steps are as follows:

[0034] According to the assessed risk level, different levels of early warning mechanisms are triggered;

[0035] When pregnant women are at low risk, continue monitoring and do not trigger an alert;

[0036] Pregnant women are advised to continue to maintain a healthy lifestyle, undergo regular checkups, and pay attention to daily self-monitoring;

[0037] When a pregnant woman is at medium risk, a primary warning is issued to remind medical staff and pregnant women to increase the frequency of monitoring;

[0038] Pregnant women are advised to adjust their diet, increase moderate exercise and / or use medication to control physiological indicators. At the same time, targeted suggestions are provided based on the individual information of pregnant women.

[0039] When a pregnant woman is at high risk, a strong warning will be issued and medical staff and the pregnant woman will be notified immediately; the pregnant woman is advised to take emergency measures immediately and develop targeted treatment plans.

[0040] In a second aspect, the present invention provides a high-risk maternal condition early warning system. Based on any step of the high-risk maternal condition early warning method as described in the first aspect of the present invention, the system includes a data acquisition module, a multi-stage risk prediction module, a fuzzy logic optimization module, a health risk assessment module and an early warning and intervention module; the data acquisition module is used to collect the individualized information and physiological data of the pregnant woman, automatically identify the pregnancy stage, and generate a maternal data packet; the multi-stage risk prediction module is used to construct a multi-stage hierarchical risk prediction model based on historical data, and predict the maternal stage health risk score according to the maternal data packet; the fuzzy logic optimization module is used to combine the maternal stage health risk score with the pregnant woman's individualized information, optimize the maternal stage health risk score through fuzzy logic, and generate a personalized maternal health risk score; the health risk assessment module is used to evaluate the current health risk level of the pregnant woman according to the personalized maternal health risk score; the early warning and intervention module is used to trigger the early warning mechanism and provide personalized intervention suggestions according to the current health risk level of the pregnant woman.

[0041] In a third aspect, an embodiment of the present invention provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the high-risk pregnant and lying-in woman condition warning method as described in the first aspect of the present invention is implemented.

[0042] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the high-risk pregnant woman's condition warning method as described in the first aspect of the present invention is implemented.

[0043] The beneficial effects of the present invention are as follows: by collecting individualized information and physiological data of pregnant women, combining historical data, building a multi-stage hierarchical risk prediction model, predicting the staged health risk score of pregnant women, and optimizing the score through fuzzy logic to generate a personalized health risk assessment. In this way, it is possible to accurately identify health risks at different stages of pregnancy, dynamically adjust the score, provide personalized early warning and intervention suggestions, and ultimately improve the accuracy of early warning of high-risk pregnant women, ensuring the health and safety of mothers and babies. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0045] Figure 1 This is a flow chart of the high-risk pregnant woman's condition early warning method in Example 1.

[0046] Figure 2 This is a block diagram of the high-risk pregnant women's condition early warning system in Example 1. DETAILED DESCRIPTION

[0047] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0048] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0049] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0050] Example 1, reference Figure 1 , which is the first embodiment of the present invention, and provides a method for early warning of a high-risk pregnant woman's condition, comprising the following steps:

[0051] S1: Collect individualized information and physiological data of pregnant women, automatically identify the pregnancy stage, and generate a data package for pregnant women.

[0052] S1.1: The individualized information of the pregnant woman includes age, medical history, family genetic history and previous pregnancy status.

[0053] It should be noted that the individualized information of pregnant women not only covers basic age factors, but also includes medical history, family genetic history and previous pregnancy conditions, which are crucial for assessing the health risks of pregnant women. For example, age is a common high-risk factor, and older pregnant women usually face a higher risk of complications; medical history and family genetic history can reveal potential genetic diseases or metabolic problems; and previous pregnancy conditions can help identify whether there is a risk of recurrent miscarriage, premature birth or other pregnancy complications.

[0054] The above information is obtained with the user's consent and is used for legal purposes.

[0055] S1.2: The physiological data include fetal heart rate, blood pressure, blood sugar, body temperature and uterine contraction intensity.

[0056] It should be noted that these indicators have important clinical significance in pregnancy monitoring. Fetal heart rate is an important parameter reflecting fetal health. Abnormal fetal heart rate may indicate fetal distress or other pregnancy complications. Changes in blood pressure (such as gestational hypertension) are early signals of high-risk diseases during pregnancy (such as preeclampsia). Monitoring of blood sugar levels is particularly important, especially for patients with gestational diabetes. Uncontrolled blood sugar may pose a serious threat to the health of mother and baby. Changes in body temperature can indicate potential infection or other health problems. The intensity of uterine contractions is one of the key indicators for assessing the risk of labor or premature birth.

[0057] S1.3: Based on the last menstrual period date of the pregnant woman, the expected date of delivery is calculated using the Naegele formula, and the current pregnancy stage of the pregnant woman is automatically identified based on the expected date of delivery and the current date.

[0058] It should be noted that in this process, the last menstrual period date of the pregnant woman is first obtained, and the expected date of delivery is calculated using the Naegele formula, which is based on the last menstrual period date, plus one year, minus three months, and then plus seven days to get the expected date of delivery. Then, the calculated expected date of delivery is compared with the current date to automatically identify the current stage of pregnancy (such as early, mid-term, late or labor period) of the pregnant woman, thereby providing an accurate time basis for subsequent risk assessment and health monitoring.

[0059] S1.4: Integrate the collected maternal physiological data and current pregnancy stage information into a maternal data package.

[0060] It should be noted that by combining the current pregnancy stage information with the real-time collected physiological data, more timely and targeted input data can be provided for the subsequent multi-stage stratified risk prediction model. This integration not only helps the model to accurately identify health risks at the current stage, but also provides basic data support for personalized early warning and intervention recommendations, ensuring the reliability and effectiveness of risk assessment.

[0061] S2: Based on historical data, a multi-stage hierarchical risk prediction model is constructed, and the maternal stage health risk score is predicted according to the maternal data package.

[0062] S2.1: Collect historical data from the hospital’s electronic health record and maternal management system database.

[0063] It should be noted that collecting historical data can provide a large amount of real clinical data support for the model, thereby improving the accuracy of risk assessment and the reliability of the prediction model.

[0064] S2.2: Label the historical data to mark the different conditions of pregnant women and the corresponding pregnancy stages.

[0065] The stages of pregnancy include: early, mid, late, and labor.

[0066] S2.3: Design independent assessment layers for early, mid, late and labor stages according to different maternal conditions;

[0067] S2.4: The early evaluation layer uses the conditional dependency inference mechanism of the Bayesian network to capture the complex associations between physiological data in the early stages;

[0068] It should be noted that in early pregnancy, the correlation between physiological data is more complex, and changes in some data may be caused by potential dependencies of other data. Bayesian network is a probabilistic graph model that can infer the mutual influence between different physiological indicators through conditional probability. The early evaluation layer uses the conditional dependency inference mechanism of Bayesian network to construct the conditional dependency relationship between multiple physiological indicators of pregnant women (such as hormone levels, weight, blood pressure, etc.), so as to infer the potential impact of certain risk factors. For example, if the Bayesian network finds through analysis that the pregnant woman's weight gain is abnormal and there is a family history of diabetes, then it can infer that the pregnant woman has a higher risk of gestational diabetes in the early stages.

[0069] S2.5: The mid-term evaluation layer uses the memory capacity of LSTM to capture the dynamic changes of physiological data over time in the mid-term stage;

[0070] It should be noted that in the second trimester, physiological data show dynamic changes over time, and these changes often have long-term dependencies. The mid-term evaluation layer uses the long short-term memory network (LSTM), which is good at processing time series data and can capture the changing patterns of physiological data at different time points. LSTM can predict future trends by memorizing data from the past period of time and help identify potential risk factors. For example, LSTM can analyze data such as blood pressure and fetal development in the second trimester, capture the trend of blood pressure gradually increasing over time, and thus predict that pregnant women may be at risk of gestational hypertension.

[0071] S2.6: The late stage assessment layer uses the feature selection and ensemble learning capabilities of XGBoost to screen out key risk factors from a large amount of physiological data in the late stage;

[0072] It should be noted that in the late pregnancy, the amount of physiological data of pregnant women increases, and the risk factors are more complex. The late assessment layer uses the XGBoost algorithm, which is an integrated learning algorithm that can screen out the most critical risk factors from a large amount of physiological data through feature selection technology. The XGBoost algorithm can prioritize the factors that have the greatest impact on late health risks through analysis of multiple physiological data (such as fetal size, blood sugar, amniotic fluid volume, etc.), and conduct key monitoring and risk prediction. For example, through the feature selection of the XGBoost algorithm, it can be found that fetal enlargement and abnormal blood sugar in pregnant women are the main risks faced in the late pregnancy, and early intervention measures can be taken for these factors.

[0073] S2.7: The labor assessment layer uses the reinforcement learning capability of Deep Q-Learning to capture the real-time dynamic changes of physiological data during the labor period;

[0074] It should be noted that during the labor period, the physiological data of pregnant women and fetuses change rapidly and timeliness requirements are very high. The labor assessment layer uses the Deep Q-Learning (deep reinforcement learning) algorithm to analyze the physiological data of pregnant women (such as uterine contraction frequency, fetal heart rate, etc.) in real time. Through the mechanism of reinforcement learning, Deep Q-Learning can continuously learn from real-time data, optimize the decision-making process, and help medical staff make rapid and accurate interventions during labor. For example, when the fetal heart rate begins to drop, the Deep Q-Learning model can quickly determine the fetal distress through real-time data feedback, and recommend that the medical team take emergency measures, such as cesarean section.

[0075] S2.8: Each independent assessment layer is independently trained with physiological data of different pregnancy stages in historical data to strengthen the perception ability of each independent assessment layer on the physiological data of pregnant women at the corresponding stage, and finally form a multi-stage hierarchical risk prediction model that can perform targeted risk prediction based on the physiological data of pregnant women at different pregnancy stages;

[0076] It should be noted that each independent assessment layer makes full use of the physiological data and specific algorithm models of that stage to ensure that the model can accurately perceive the physiological changes at that stage and capture potential risks in a targeted manner. In addition, the models of each assessment layer do not exist in isolation, but work together. Although the data of each stage is trained independently, the output of the final model is a comprehensive risk assessment result, which requires the combination and interpretation of the independent assessment results of each stage through the subsequent integration layer. Specifically, information can be transmitted between assessment layers through a feedforward mechanism.

[0077] For example, the results of the early evaluation layer can be used as one of the initial inputs of the mid-term evaluation layer to help LSTM better capture the dynamics of the mid-term time series. The late evaluation layer can further screen out key risk factors by combining the potential risk information provided by the previous evaluation layer. The labor assessment layer needs to use the accumulated risk information of the previous stages and combine it with the real-time dynamic feedback of the reinforcement learning algorithm to make more accurate real-time decisions.

[0078] S2.9: Input the maternal data packet into the multi-stage hierarchical risk assessment model, and dynamically select the assessment layer for risk score prediction based on the current expected delivery stage of the pregnant woman. The expression is: ;in, Indicates the health risk score of the pregnant woman in the current pregnancy stage. represents the physiological data of pregnant women at time t, represents the hyperbolic tangent function, represents the influence of the ith physiological data on the risk score, represents the mean of the ith physiological data, represents the standard deviation of the ith physiological data, and n is the total number of physiological data.

[0079] It should be noted that the core of this expression is the hyperbolic tangent function. Physiological data of the current stage of pregnancy and parturition The mean of predefined physiological data Specifically, each physiological index is first calculated With its mean The standardized difference between Normalized), and then according to the weights (i.e. ) to weight the squares of these differences. Next, use The function compresses the result to the interval [-1,1], and then converts the output into a health risk score in the interval [0,1] by translation and scaling. , to reflect the health status of pregnant women at time t.

[0080] S3: Combine the maternal stage health risk score with the individualized information of the pregnant women, optimize the maternal stage health risk score through fuzzy logic, and generate a personalized maternal health risk score.

[0081] S3.1: Select the individual information of pregnant women and the health risk score of pregnant women in the current pregnancy stage through fuzzy processing As input variables, fuzzy sets are divided for each input variable and membership functions are defined. The precise values ​​of the input variables are mapped to the fuzzy sets through membership functions to generate fuzzy information.

[0082] For example, age is converted into "young", "middle-aged" or "old", and blood sugar level is converted into fuzzy categories such as "normal" or "high". These fuzzy values ​​are based on a certain degree of membership, indicating the fuzziness of different states of pregnant women.

[0083] S3.2: Based on clinical experience and historical data, select a health risk score that is appropriate for the current stage of pregnancy. Relevant individualized information of pregnant women, by defining individualized information of pregnant women and health risk scores of pregnant women in the current stage of pregnancy The causal relationship is established by setting fuzzy rules;

[0084] For example, rule 1: if the age is "old" and the risk is "medium", then increase the health risk score.

[0085] Rule 2: If blood sugar is “high”, increase the risk score.

[0086] Rule 3: If there is no high-risk medical history and the age is "young", maintain or reduce the risk score;

[0087] S3.3: Match the fuzzified information with the fuzzy rules, calculate the activation degree of each rule through the Mamdani algorithm and synthesize the fuzzy output;

[0088] It should be noted that the activation degree of each rule is calculated as follows: ;in, is the activation degree of the jth fuzzy rule, is the input variable For fuzzy sets The membership degree of is the input variable For fuzzy sets The membership degree of is the input variable For fuzzy sets The degree of membership;

[0089] The synthetic fuzzy output is expressed as: ;in, is the fuzzy output result after synthesis, Represents output variable In the fuzzy set The degree of membership in .

[0090] S3.4: The synthesized fuzzy output is converted into a numerical health risk score through defuzzification, and then the health risk score of the pregnant woman in the current pregnancy stage is scored. Dynamically adjust to generate personalized maternal health risk scores .

[0091] For example, if the initial risk score is 0.4, if the blood sugar is high, according to the fuzzy rule, Increase, if there is a high-risk history, further increase according to fuzzy rules ,Finally, the risk score was adjusted from ,0.4 to 0.6.

[0092] S4: Assess the current health risk level of pregnant women based on the personalized maternal health risk score.

[0093] S4.1: Based on historical data, define a low score threshold T1 and a high score threshold T2 to assess the current health risk level of pregnant women;

[0094] when When ≤T1, the current maternal health risk is considered to be low;

[0095] When T1< When ≤T2, the current maternal health is considered to be at medium risk;

[0096] when >T2, the current maternal health is considered to be at high risk.

[0097] It should be noted that by analyzing a large amount of historical data and extracting the statistical distribution of health risk scores, these two thresholds can be reasonably set. When it is lower than T1, it is judged as low risk; when it is between T1 and T2, it is judged as medium risk; when it is higher than T2, it is high risk. This risk grading method ensures the scientificity and reliability of the assessment, thus providing an accurate basis for subsequent early warning and intervention measures.

[0098] S5: Trigger the early warning mechanism and provide personalized intervention recommendations based on the current health risk level of the pregnant woman.

[0099] According to the assessed risk level, different levels of early warning mechanisms are triggered;

[0100] When pregnant women are at low risk, continue monitoring and do not trigger an alert;

[0101] Pregnant women are advised to continue to maintain a healthy lifestyle, undergo regular checkups, and pay attention to daily self-monitoring;

[0102] When a pregnant woman is at medium risk, a primary warning is issued to remind medical staff and pregnant women to increase the frequency of monitoring; for example, it may be recommended to increase the number of examinations, such as routine prenatal examinations once a week or additional monitoring of specific physiological indicators (such as blood pressure, blood sugar);

[0103] Pregnant women are advised to adjust their diet, increase moderate exercise and / or use medication to control physiological indicators. At the same time, targeted advice is provided based on the individualized information of the pregnant women (such as medical history); for example, if the pregnant woman has a history of gestational diabetes, a stricter blood sugar control plan is recommended, or the frequency of fetal development monitoring is increased.

[0104] When a pregnant woman is at high risk, a strong warning is issued to immediately notify medical staff and the pregnant woman; for example, go to the hospital immediately for a detailed examination, or arrange hospitalization for observation to prevent potential serious complications.

[0105] It is recommended that pregnant women take emergency measures immediately, such as hospitalization for observation, detailed fetal heart monitoring, and multidisciplinary consultation (obstetrics, internal medicine, pediatrics, etc.) to develop targeted treatment plans; for example, by adjusting diet and medication to control blood sugar, using safe antihypertensive drugs to manage hypertension, increasing the frequency of fetal heart monitoring, intervention measures to prevent premature birth, and developing personalized delivery plans for high-risk pregnant women to ensure the health and safety of mother and child.

[0106] It should also be noted that the relevant personnel were notified through various channels as follows:

[0107] Pregnant women: Send early warning information and intervention suggestions to pregnant women through mobile phone applications, text messages or phone calls, and provide relevant health education resources.

[0108] Medical staff: Send detailed health reports and early warning notifications to the responsible doctor or nurse to ensure timely understanding of the health status of pregnant women.

[0109] See also Figure 2 ,This embodiment also provides a high-risk maternal condition early warning system, including: a data acquisition module, a multi-stage risk prediction module, a fuzzy logic optimization module, a health risk assessment module and an early warning and intervention module;

[0110] Data collection module, used to collect individual information and physiological data of pregnant women, automatically identify the pregnancy stage, and generate data packets for pregnant women;

[0111] The multi-stage risk prediction module is used to build a multi-stage hierarchical risk prediction model based on historical data and predict the maternal health risk score according to the maternal data package;

[0112] The fuzzy logic optimization module is used to combine the maternal health risk score at each stage with the individual information of the pregnant women, optimize the maternal health risk score at each stage through fuzzy logic, and generate a personalized maternal health risk score;

[0113] The health risk assessment module is used to assess the current health risk level of pregnant women based on personalized maternal health risk scores;

[0114] The early warning and intervention module is used to trigger the early warning mechanism and provide personalized intervention suggestions based on the current health risk level of pregnant women.

[0115] This embodiment also provides a computer device, which is suitable for the case of a high-risk pregnant woman's condition warning method, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the high-risk pregnant woman's condition warning method as proposed in the above embodiment.

[0116] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covered on the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0117] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for realizing high-risk pregnant women's condition warning as proposed in the above embodiment is implemented; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0118] In summary, the present invention collects individualized information and physiological data of pregnant women, combines historical data, constructs a multi-stage hierarchical risk prediction model, predicts the staged health risk score of pregnant women, and optimizes the score through fuzzy logic to generate a personalized health risk assessment. In this way, the system can accurately identify health risks at different stages of pregnancy, dynamically adjust the score, provide personalized early warning and intervention suggestions, and ultimately improve the accuracy of early warning of high-risk pregnant women and ensure the health and safety of mothers and babies.

[0119] 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 preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for early warning of high-risk maternal conditions, characterized in that: include: Collect individualized information and physiological data of pregnant women, automatically identify the pregnancy stage, and generate data packages for pregnant women; Based on historical data, a multi-stage hierarchical risk prediction model is constructed, and the maternal health risk score at each stage is predicted according to the maternal data package; Combine the maternal health risk scores at different stages with the individual information of pregnant women, optimize the maternal health risk scores at different stages through fuzzy logic, and generate personalized maternal health risk scores; Assess the current health risk level of pregnant women based on personalized maternal health risk scores; According to the current health risk level of pregnant women, trigger the early warning mechanism and provide personalized intervention suggestions; The individualized information of the pregnant woman includes age, medical history, family genetic history and previous pregnancy status; The physiological data include fetal heart rate, blood pressure, blood sugar, body temperature, uterine contraction intensity; The specific steps of automatically identifying the pregnancy stage and generating a data packet for pregnant women are as follows: Based on the last menstrual period date of the pregnant woman, the expected date of delivery is calculated using the Naegele formula, and the current stage of pregnancy of the pregnant woman is identified based on the expected date of delivery and the current date; Integrate the collected maternal physiological data and current pregnancy stage information into a maternal data package; Based on the historical data, a multi-stage hierarchical risk prediction model is constructed, and the maternal health risk score is predicted according to the maternal data package. The specific steps are as follows: Historical data were collected from the hospital's electronic health record and maternal management system database; Label the historical data to mark the different conditions of pregnant women and the corresponding pregnancy stages; According to the different stages of pregnancy corresponding to different conditions of pregnant women, independent assessment layers are designed for early, mid-term, late and labor periods; The early evaluation layer uses the conditional dependency inference mechanism of the Bayesian network to capture the complex associations between physiological data in the early stages; The mid-term evaluation layer uses the memory capacity of LSTM to capture the dynamic changes of physiological data over time in the mid-term stage; The late-stage evaluation layer uses XGBoost’s feature selection and ensemble learning capabilities to screen out key risk factors from a large amount of physiological data in the late stage; The labor assessment layer uses the reinforcement learning capabilities of Deep Q-Learning to capture the real-time dynamic changes of physiological data during the labor stage; Each independent assessment layer is independently trained with physiological data from different pregnancy stages in historical data to strengthen the perception of each independent assessment layer on the physiological data of pregnant women at the corresponding stage, and ultimately forms a multi-stage hierarchical risk prediction model that can perform targeted risk prediction based on the physiological data of pregnant women at different pregnancy stages; The maternal data package is input into the multi-stage hierarchical risk assessment model. According to the current expected delivery stage of the pregnant woman, the assessment layer is dynamically selected for risk score prediction. The expression is: ； in, Indicates the health risk score of the pregnant woman in the current pregnancy stage. represents the physiological data of pregnant women at time t, represents the hyperbolic tangent function, represents the influence of the ith physiological data on the risk score, represents the mean value of the ith physiological data, represents the standard deviation of the ith physiological data, and n is the total number of physiological data.

2. The high-risk maternal condition early warning method according to claim 1, characterized in that: The maternal stage health risk score is combined with the individualized information of the pregnant woman, and the maternal stage health risk score is optimized through fuzzy logic to generate a personalized maternal health risk score. The specific steps are as follows: Through fuzzy processing, the individual information of pregnant women and the health risk score of pregnant women in the current pregnancy stage are selected. As input variables, fuzzy sets are divided for each input variable and membership functions are defined. The precise values ​​of the input variables are mapped to the fuzzy sets through membership functions to generate fuzzy information. Based on clinical experience and historical data, select a health risk score that is relevant to the current stage of pregnancy. Relevant individualized information of pregnant women, by defining individualized information of pregnant women and health risk scores of pregnant women in the current stage of pregnancy The causal relationship is established by setting fuzzy rules; Match the fuzzified information with the fuzzy rules, calculate the activation degree of each rule through the Mamdani algorithm and synthesize the fuzzy output; The synthesized fuzzy output is converted into a numerical health risk score through defuzzification, and then the health risk score of the pregnant woman in the current pregnancy stage is scored. Dynamically adjust to generate personalized maternal health risk scores .

3. The high-risk maternal condition early warning method according to claim 2, characterized in that: The specific steps of evaluating the current health risk level of pregnant women according to the personalized maternal health risk score are as follows: Based on historical data, a low score threshold T1 and a high score threshold T2 are defined to assess the current health risk level of pregnant women. when When ≤T1, the current maternal health risk is considered to be low; When T1< When ≤T2, the current maternal health is considered to be at medium risk; when >T2, the current maternal health is considered to be at high risk.

4. The high-risk maternal condition early warning method according to claim 3, characterized in that: According to the current health risk level of pregnant women, the early warning mechanism is triggered and personalized intervention suggestions are provided. The specific steps are as follows: According to the assessed risk level, different levels of early warning mechanisms are triggered; When pregnant women are at low risk, continue monitoring and do not trigger an alert; Pregnant women are advised to continue to maintain a healthy lifestyle, undergo regular checkups, and pay attention to daily self-monitoring; When a pregnant woman is at medium risk, a primary warning is issued to remind medical staff and pregnant women to increase the frequency of monitoring; Pregnant women are advised to adjust their diet, increase moderate exercise and / or use medication to control physiological indicators. At the same time, targeted suggestions are provided based on the individual information of pregnant women. When pregnant women are at high risk, a strong warning is issued to immediately notify medical staff and pregnant women; It is recommended that pregnant women take emergency measures immediately and develop targeted treatment plans.

5. A high-risk maternal condition early warning system, based on the high-risk maternal condition early warning method according to any one of claims 1 to 4, characterized in that: include: Data collection module, multi-stage risk prediction module, fuzzy logic optimization module, health risk assessment module and early warning and intervention module; Data collection module, used to collect individual information and physiological data of pregnant women, automatically identify the pregnancy stage, and generate data packets for pregnant women; The multi-stage risk prediction module is used to build a multi-stage hierarchical risk prediction model based on historical data and predict the maternal health risk score according to the maternal data package; The fuzzy logic optimization module is used to combine the maternal health risk score at each stage with the individual information of the pregnant women, optimize the maternal health risk score at each stage through fuzzy logic, and generate a personalized maternal health risk score; The health risk assessment module is used to assess the current health risk level of pregnant women based on personalized maternal health risk scores; The early warning and intervention module is used to trigger the early warning mechanism and provide personalized intervention suggestions based on the current health risk level of pregnant women.

6. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the high-risk pregnant woman's condition warning method according to any one of claims 1 to 4 are implemented.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the high-risk pregnant woman's condition early warning method according to any one of claims 1 to 4 are implemented.

Images