A real-time obstetric nursing optimization method based on big data

Abstract

The present invention discloses a real-time obstetric care optimization method based on big data, which relates to the field of artificial intelligence technology, including: real-time collection of maternal and fetal physiological data, and pre-processing of the maternal and fetal physiological data by fusing them with historical health data to form a multi-dimensional health data set; constructing a personalized dynamic health model, analyzing the physiological changes of the mother and fetus in real time, and predicting future health trends; fusing the predicted future health trends with fetal electrocardiogram, maternal electrocardiogram and environmental data to identify potential health risks; classifying abnormal events according to risk assessment results, generating personalized care plans, and automatically optimizing the generated personalized care plans; through the intelligent feedback mechanism of the present invention, the system can generate and dynamically optimize personalized care plans, adjust the care path in real time, and significantly improve the accuracy and reliability of obstetric care.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a real-time obstetric nursing optimization method based on big data. Background Art

[0002] With the rapid development of big data technology, the Internet of Things and artificial intelligence, the medical field, especially obstetric care, is gradually shifting towards intelligence and personalization. Traditional obstetric care mainly relies on regular clinical examinations and doctors' experience and judgment, and cannot monitor pregnant women and fetuses all the time, especially in the late pregnancy, when the physical condition of pregnant women changes frequently and the health of the fetus is also affected by many factors.

[0003] However, existing technical means can usually only provide single-dimensional physiological data monitoring, and fail to fully utilize historical health data and multimodal data for comprehensive health trend analysis. In addition, existing systems often rely on preset rules in terms of abnormality detection and formulation of care plans, lacking the ability to dynamically adjust and personalize, and are unable to cope with complex and changing clinical situations. Summary of the invention

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

[0005] Therefore, the present invention provides a real-time obstetric care optimization method based on big data to solve the problem that traditional obstetric care cannot integrate multi-dimensional data in real time, predict health trends, and dynamically optimize personalized care plans.

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

[0007] In a first aspect, an embodiment of the present invention provides a method for optimizing real-time obstetric care based on big data, comprising: collecting physiological data of the mother and the fetus in real time, and fusing and preprocessing the physiological data of the mother and the fetus with historical health data to form a multi-dimensional health data set;

[0008] Build a personalized dynamic health model to analyze the physiological changes of mothers and fetuses in real time and predict future health trends;

[0009] Fusion of predicted future health trends with fetal ECG, maternal ECG and environmental data to identify potential health risks;

[0010] Abnormal events are classified according to the risk assessment results, personalized care plans are generated, and the generated personalized care plans are automatically optimized.

[0011] As a preferred embodiment of the obstetric real-time nursing optimization method based on big data of the present invention, the maternal physiological data includes heart rate, blood pressure, blood sugar, body temperature, respiratory rate and blood oxygen saturation;

[0012] The fetal physiological data includes fetal electrocardiogram signal, fetal movement frequency and uterine contraction strength.

[0013] As a preferred solution of the obstetric real-time nursing optimization method based on big data described in the present invention, the physiological data of the mother and the fetus are fused and preprocessed with the historical health data to form a multi-dimensional health data set, including the following steps:

[0014] Build a data integration platform to clean, convert and synchronize maternal and fetal physiological data from different devices;

[0015] Combine the processed data with historical health records to form a multi-dimensional health dataset.

[0016] As a preferred solution of the obstetric real-time nursing optimization method based on big data described in the present invention, building a personalized dynamic health model, analyzing the physiological changes of the mother and the fetus in real time, and predicting future health trends include the following steps:

[0017] Input the physiological data of the mother and fetus into the long short-term memory network in deep learning, extract the time series features by analyzing the trend of physiological data changing over time, extract the frequency domain features by analyzing the frequency components of physiological signals, extract the abnormal features by analyzing the fluctuation amplitude and change rate of physiological parameters, and extract the correlation features by capturing the dependency relationship between the mother and the fetus;

[0018] According to the analysis results, the future health trend is predicted, and the expression is:

[0019] ;

[0020] in, It's time The health prediction value of It is Physiological data at the moment The value of It is The mean of physiological data, It is The standard deviation of physiological data, is the physiological data index, It is The weight of physiological data, yes Health status at all times, is the current time, is the time constant, is the amount of physiological data.

[0021] As a preferred solution of the obstetric real-time nursing optimization method based on big data described in the present invention, the predicted future health trend is integrated with the fetal ECG, maternal ECG and environmental data to identify potential health risks, including the following steps:

[0022] Collect fetal ECG, maternal ECG and environmental data in real time, and extract features from each modality data;

[0023] The weight is dynamically adjusted according to the contribution of each modal data to the health status in the past period of time. The expression is:

[0024] ;

[0025] in, It is The weight of a specific modality data, is the modality-specific data index, is the number of specific modes, It is The historical contribution of a specific modality data, is the contribution of traversing all modal data, It is The historical contribution of a specific modality of data;

[0026] The data of each modality are weighted and fused to obtain multimodal comprehensive data, which is expressed as:

[0027] ;

[0028] in, It's time Multimodal comprehensive data, It's time No. The actual value of the modal data, is the modal data index, is the number of modal data;

[0029] Integrate variational autoencoders and convolutional neural networks to capture the global and local features of each modality data;

[0030] Use variational autoencoders to learn historical normal data, generate potential distributions for each modal data, encode real-time input modal data based on variational autoencoders, and compare them with the potential distributions of historical normal data to calculate the degree of deviation;

[0031] Use convolutional neural networks to extract local features from modal data and detect whether there are local abnormal signals;

[0032] The anomaly score of the current multimodal comprehensive data is calculated through the anomaly scoring mechanism based on KL divergence. The expression is:

[0033] ;

[0034] in, It is at the moment The abnormal score of It's time Multimodal comprehensive data, It's time The current distribution of multimodal synthetic data It is The normal distribution of the comprehensive data calculated from multimodal data, It's time The current distribution of health prediction values, It's time The health prediction value of It is Normal distribution of health prediction values ​​for modal data, is the time factor, is an asymmetric measure of the difference between two probability distributions.

[0035] As a preferred solution of the obstetric real-time nursing optimization method based on big data described in the present invention, classifying abnormal events according to risk assessment results and generating a personalized nursing plan include the following steps:

[0036] Abnormal events are classified into low risk, medium risk and high risk according to the size of the abnormality score;

[0037] When low risk is detected, no immediate intervention is required and personalized advice is provided to pregnant women;

[0038] When a moderate risk is detected, a notification is automatically sent to the doctor to provide the current abnormal details, and the doctor is advised to observe and further monitor relevant physiological data through the remote monitoring platform, provide nursing guidance, push abnormality reports to pregnant women, and recommend regular clinical examinations;

[0039] When a high-risk abnormality is detected, the doctor is notified immediately and the pregnant woman is advised to go to the hospital immediately. The frequency of fetal heart monitoring is automatically increased, the physiological signals of the fetus and the mother are tracked in real time, an alarm is sent through the emergency notification platform, and an emergency care plan is initiated.

[0040] As a preferred solution of the obstetric real-time nursing optimization method based on big data described in the present invention, automatically optimizing the generated personalized nursing plan includes the following steps:

[0041] Regularly collect new multimodal data and user feedback;

[0042] Evaluate the effectiveness of current care programs based on new data and user feedback;

[0043] Based on the evaluation results, personalized suggestions are added and the user experience is improved to form an optimized care plan.

[0044] In a second aspect, an embodiment of the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, any step of the real-time obstetric care optimization method based on big data as described in the first aspect of the present invention is implemented.

[0045] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any step of the real-time obstetric care optimization method based on big data as described in the first aspect of the present invention.

[0046] The beneficial effects of the present invention are as follows: by collecting the physiological data of the mother and fetus in real time, and fusing and preprocessing it with historical health data, a multi-dimensional health data set is formed, thereby realizing all-round monitoring of pregnant women and fetuses; a personalized dynamic health model is constructed to analyze physiological changes in real time and predict health trends, thereby improving the early warning capability of the system; the prediction results are integrated with the fetal ECG, maternal ECG and environmental data, and a multimodal adaptive weighted fusion method and a convolutional neural network are used to identify potential health risks and classify abnormal events. The system automatically selects a feedback mechanism according to the risk level and takes corresponding intervention measures to ensure the timeliness and effectiveness of care. Through the intelligent feedback mechanism, the system can generate and dynamically optimize personalized care plans, adjust the care path in real time, significantly improve the accuracy and reliability of obstetric care, reduce potential health risks, and ensure the safety of mothers and babies. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] 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.

[0048] Figure 1 This is a flow chart of the obstetric real-time care optimization method based on big data in Example 1.

[0049] Figure 2 This is a flow chart of the abnormality scoring in Example 1. DETAILED DESCRIPTION

[0050] 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.

[0051] 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.

[0052] 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.

[0053] Example 1, reference Figure 1 and Figure 2 , which is the first embodiment of the present invention, and provides a real-time obstetric nursing optimization method based on big data, comprising the following steps:

[0054] S1, real-time collection of maternal and fetal physiological data, and fusion preprocessing of maternal and fetal physiological data with historical health data to form a multi-dimensional health data set, including the following steps:

[0055] The mother's physiological data includes heart rate, blood pressure, blood sugar, body temperature, respiratory rate and blood oxygen saturation; the fetus's physiological data includes fetal electrocardiogram signals, fetal movement frequency and uterine contraction intensity; build a data integration platform to clean, convert and synchronize timestamp data from different devices; combine the processed data with historical health records to form a multi-dimensional health data set.

[0056] It should be noted that by collecting a variety of physiological data of the mother and fetus, including heart rate, blood pressure, blood sugar, body temperature, respiratory rate, blood oxygen saturation, fetal ECG signal, fetal movement frequency and uterine contraction strength, a comprehensive monitoring of the health status of the mother and fetus is achieved. For example, heart rate and blood pressure can reflect the health status of the cardiovascular system, blood sugar can reflect the metabolic state, body temperature and blood oxygen saturation can reflect the overall health status of the body, fetal ECG signal and fetal movement frequency can reflect the health status of the fetus, and uterine contraction strength can reflect the preparation before delivery. Through multi-dimensional data collection, potential health problems can be discovered early, and intervention measures can be taken in time to reduce maternal and child health risks; by integrating historical health data, personalized health assessment can be achieved. For example, if a pregnant woman has a history of hypertension, the system can pay more attention to changes in blood pressure data and discover potential hypertension risks in time. Genetic data can provide genetic information to help identify the risk of genetic diseases; the integration of historical health data enables the system to conduct long-term trend analysis and discover changing trends in health status. For example, by analyzing past pregnancy health records, we can find the long-term change patterns of certain physiological parameters, which can provide a basis for future health management and prevention; through data smoothing, we can eliminate short-term fluctuations and noise in the data, making the data more stable and reliable. This helps to reduce false positives and false negatives and improve the accuracy of health assessment. For example, heart rate data may be affected by short-term activities. Through smoothing, these short-term fluctuations can be filtered out to better reflect the true physiological state; noise reduction can remove interference signals in the data and improve the quality of the data. This is especially important for physiological data, because physiological signals are often interfered by many factors, such as environmental noise, equipment noise, etc. Noise reduction can improve the signal-to-noise ratio of the data and make subsequent analysis more accurate; standardization can convert different types of physiological data to the same dimension for comparison and analysis. This helps to eliminate the dimensional differences between different data and improve the consistency and comparability of the data. For example, the data units of heart rate and blood pressure are different. Through standardization, they can be converted to the same dimension for comprehensive analysis.

[0057] S2, builds a personalized dynamic health model to analyze the physiological changes of the mother and fetus in real time and predict future health trends, including the following steps:

[0058] Input the physiological data of the mother and fetus into the long short-term memory network in deep learning, extract the time series features by analyzing the trend of physiological data changing over time, extract the frequency domain features by analyzing the frequency components of physiological signals, extract the abnormal features by analyzing the fluctuation amplitude and change rate of physiological parameters, and extract the correlation features by capturing the dependency relationship between the mother and the fetus;

[0059] According to the analysis results, the future health trend is predicted, and the expression is:

[0060] ;

[0061] in, It's time The health prediction value of It is Physiological data at the moment values, including heart rate ,blood pressure , fetal movement frequency , fetal ECG signal and uterine contraction strength , It is The mean of physiological data, It is The standard deviation of physiological data, is the physiological data index, It is The weights of the physiological data are set according to their impact on overall health. The weights are set as follows: Fetal ECG signal =0.3, uterine contraction intensity =0.25, fetal movement frequency = 0.2, maternal heart rate =0.15, maternal blood pressure =0.1, yes yes The health status at the moment, the value range is 0≤ ≤1, is the current time, is a time constant used to adjust the decay rate of the influence of historical data. It is the quantity of physiological data. According to the specific situation, the types of physiological data used can be flexibly adjusted to more comprehensively capture the health status of the mother and fetus.

[0062] Furthermore, The range of is real number, when When >0, it indicates good health. The larger the value, the more stable the physiological state of the mother and fetus. ≈0, indicating that the health status is critical and requires further observation and monitoring; When <0, it indicates that there is a health risk. The smaller the value, the greater the risk, and the system should issue an early warning in a timely manner.

[0063] It should be noted that by using the long short-term memory network (LSTM) in deep learning to analyze the multidimensional data set, the key features are extracted, and the physiological changes of the mother and fetus are accurately captured. The LSTM network can effectively process time series data and capture long-term dependencies, thereby extracting the most representative features from complex physiological data. This step not only improves the efficiency of data processing, but also provides high-quality input data for subsequent health prediction models, ensuring the accuracy and reliability of predictions. Finally, by extracting key features, the system can more accurately identify potential health risks and provide a scientific basis for early intervention; by building a personalized health prediction model, the system can predict future health trends based on the analysis results. This model is based on the time series modeling capability of the LSTM network, which can capture the dynamic changes in maternal and fetal physiological data and predict future health status. The establishment of a personalized prediction model enables each pregnant woman's health status to be evaluated and predicted in a targeted manner, improving the accuracy and applicability of the prediction. In addition, by predicting future health trends, the system can issue early warnings before potential health problems occur, providing doctors and pregnant women with a valuable time window, so that timely intervention measures can be taken to effectively prevent the occurrence of health risks.

[0064] S3, combines the predicted future health trends with fetal ECG, maternal ECG and environmental data to identify potential health risks, including the following steps:

[0065] Collect fetal ECG, maternal ECG and environmental data in real time, and extract features from each modality data;

[0066] The weight is dynamically adjusted according to the contribution of each modal data to the health status in the past period of time. The expression is:

[0067] ;

[0068] in, It is The weight of a specific modality data, is the modality-specific data index, is the number of specific modes, It is The historical contribution of a specific modality data, is the contribution of traversing all modal data, It is The historical contribution of a specific modality of data;

[0069] The data of each modality are weighted and fused to obtain multimodal comprehensive data, which is expressed as:

[0070] ;

[0071] in, It's time Multimodal comprehensive data, It's time No. The actual value of the modal data, is the modal data index, is the number of modal data;

[0072] A collection model based on variational autoencoders and convolutional neural networks is used to capture the global and local features of each modality data at the same time;

[0073] Use variational autoencoders to learn historical normal data, generate potential distributions for each modal data, encode real-time input modal data based on variational autoencoders, and compare them with the potential distributions of historical normal data to calculate the degree of deviation;

[0074] Use convolutional neural networks to extract local features from modal data and detect whether there are local abnormal signals (such as sudden abnormalities in maternal ECG signals, abnormal fluctuations in fetal ECG signals, etc.);

[0075] The anomaly score of the current multimodal comprehensive data is calculated through the anomaly scoring mechanism based on KL divergence. The expression is:

[0076] ;

[0077] in, It is at the moment The abnormal score of It's time Multimodal comprehensive data, It's time The current distribution of multimodal synthetic data, It is The normal distribution of the comprehensive data calculated from multimodal data, is the current distribution of predicted health values ​​at time 𝑡, It's time The health prediction value of It is Normal distribution of health prediction values ​​for modal data, is the time factor, is an asymmetric measure of the difference between two probability distributions.

[0078] It should be noted that wearable devices and medical-grade monitoring instruments collect fetal electrocardiogram (FECG), maternal electrocardiogram (MECG) and environmental data (such as temperature, humidity, noise level) in real time. Feature extraction of each modal data, such as heart rate variability, RR interval, blood pressure fluctuation, etc., can timely reflect the physiological changes of the mother and fetus and improve the sensitivity and accuracy of monitoring. Feature extraction helps to extract key information from complex data, facilitate subsequent analysis and modeling, quickly discover potential health problems, improve the ability of early warning, and thus take timely intervention measures to ensure the safety of mother and child; by dynamically adjusting weights, the model can better capture the impact of different modal data on health status and improve the accuracy and reliability of health risk identification; modal data fusion can comprehensively utilize information from different sources, improve the comprehensiveness and comprehensiveness of the model, and better reflect the overall health status of the mother and fetus; through multimodal data fusion and deep learning models, accurate identification and classification of potential health risks can be achieved, the sensitivity and accuracy of early warning can be improved, and intervention measures can be taken in time.

[0079] S4, classifying abnormal events according to the risk assessment results, generating personalized care plans, and automatically optimizing the generated personalized care plans, including the following steps:

[0080] Abnormal events are classified into low risk, medium risk and high risk according to the size of the abnormality score;

[0081] When low risk is detected, no immediate intervention is required and personalized advice is provided to pregnant women;

[0082] When a moderate risk is detected, a notification is automatically sent to the doctor to provide the current abnormal details, and the doctor is advised to observe and further monitor relevant physiological data through the remote monitoring platform, provide nursing guidance, push abnormality reports to pregnant women, and recommend regular clinical examinations;

[0083] When a high-risk abnormality is detected, the doctor is notified immediately and the pregnant woman is advised to go to the hospital immediately. The frequency of fetal heart monitoring is automatically increased, the physiological signals of the fetus and the mother are tracked in real time, an alarm is sent through the emergency notification platform, and an emergency care plan is initiated;

[0084] Regularly collect new multimodal data and user feedback;

[0085] Evaluate the effectiveness of current care programs based on new data and user feedback;

[0086] Based on the evaluation results, personalized suggestions are added and the user experience is improved to form an optimized care plan.

[0087] Furthermore, the low risk threshold for abnormal classification is set to , the high risk threshold is , the classification standards for abnormal events are as follows:

[0088] When 0< ≤ When , it means the abnormality score is low and the risk is low; when < ≤ When , it indicates a medium risk; when > , it indicates a high risk.

[0089] It should be noted that by calculating the abnormality score, abnormal events are divided into low risk, medium risk and high risk, so that the system can take different intervention measures according to different levels of risk, avoiding a one-size-fits-all management approach, improving the accuracy and effectiveness of nursing, ensuring the rational allocation of resources, avoiding excessive intervention in low-risk situations, timely monitoring and guidance in medium-risk situations, and taking emergency measures quickly in high-risk situations, thereby improving the efficiency and safety of overall nursing; generating personalized health recommendations based on the pregnant woman’s physiological data and historical health records, such as a reasonable diet and moderate exercise, to help pregnant women better manage their health in daily life, prevent potential health problems, enhance the self-management ability of pregnant women, reduce unnecessary medical interventions, and improve the quality of life and satisfaction of pregnant women; doctors can provide remote consultation and consultation to pregnant women, so as to improve their quality of life and satisfaction. The process monitoring platform monitors the physiological data of pregnant women and fetuses in real time, provides professional nursing guidance, and pushes abnormal reports to pregnant women. It recommends that pregnant women undergo regular clinical examinations, ensuring that doctors can understand the health status of pregnant women in a timely manner and take necessary monitoring and guidance measures to avoid further deterioration of risks, improve doctors' response speed and processing efficiency, reduce the possibility of medium risk turning into high risk, and ensure the safety of pregnant women and fetuses; automatically increase the frequency of fetal heart monitoring, track the physiological signals of the fetus and mother in real time, and send alarms through the emergency notification platform to initiate emergency care plans, ensuring that in high-risk situations, emergency measures can be taken quickly to deal with potential health crises in a timely manner. Through rapid response and emergency treatment, the harm of high-risk events to pregnant women and fetuses is minimized, and the success rate of first aid is improved.

[0090] This embodiment also provides a computer device, which is suitable for the obstetric real-time care optimization method based on big data, 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 obstetric real-time care optimization method based on big data proposed in the above embodiment.

[0091] 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.

[0092] 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 optimizing obstetric real-time nursing based on big data 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 (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, disk or optical disk.

[0093] In summary, the present invention collects the physiological data of the mother and fetus in real time, and integrates and pre-processes it with historical health data to form a multi-dimensional health data set, thereby realizing all-round monitoring of pregnant women and fetuses; by constructing a personalized dynamic health model, it analyzes physiological changes in real time and predicts health trends, thereby improving the early warning ability of the system; the prediction results are integrated with the fetal ECG, maternal ECG and environmental data, and a multimodal adaptive weighted fusion method and convolutional neural network are used to identify potential health risks and classify abnormal events. The system automatically selects a feedback mechanism according to the risk level and takes corresponding intervention measures to ensure the timeliness and effectiveness of care. Through the intelligent feedback mechanism, the system can generate and dynamically optimize personalized care plans, adjust the care path in real time, significantly improve the accuracy and reliability of obstetric care, reduce potential health risks, and ensure the safety of mothers and babies.

[0094] 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 real-time obstetric nursing optimization method based on big data, characterized in that: include: Real-time collection of maternal and fetal physiological data, and fusion and preprocessing of maternal and fetal physiological data with historical health data to form a multi-dimensional health data set; Build a personalized dynamic health model to analyze the physiological changes of mothers and fetuses in real time and predict future health trends; Fusion of predicted future health trends with fetal ECG, maternal ECG and environmental data to identify potential health risks; Classify abnormal events according to risk assessment results, generate personalized care plans, and automatically optimize the generated personalized care plans; The maternal physiological data includes heart rate, blood pressure, blood sugar, body temperature, respiratory rate and blood oxygen saturation; The fetal physiological data includes fetal electrocardiogram signal, fetal movement frequency and uterine contraction strength; The following steps are involved in fusing the physiological data of the mother and fetus with the historical health data to form a multi-dimensional health data set: Build a data integration platform to clean, convert and synchronize maternal and fetal physiological data from different devices; Combine the processed data with historical health records to form a multi-dimensional health data set; Building a personalized dynamic health model to analyze the physiological changes of the mother and fetus in real time and predict future health trends includes the following steps: Input the physiological data of the mother and fetus into the long short-term memory network in deep learning, extract the time series features by analyzing the trend of physiological data changing over time, extract the frequency domain features by analyzing the frequency components of physiological signals, extract the abnormal features by analyzing the fluctuation amplitude and change rate of physiological parameters, and capture the dependency relationship between the mother and the fetus to extract the correlation features; According to the analysis results, the future health trend is predicted, and the expression is: ； in, It's time The health prediction value of It is The value of physiological data at time 𝑡, It is The mean of physiological data, It is The standard deviation of physiological data, is the physiological data index, It is The weight of physiological data, yes Health status at all times, is the current time, is the time constant, is the amount of physiological data; The following steps are involved in integrating the predicted future health trends with the fetal ECG, maternal ECG and environmental data to identify potential health risks: Collect fetal ECG, maternal ECG and environmental data in real time, and extract features from each modality data; The weight is dynamically adjusted according to the contribution of each modal data to the health status in the past period of time. The expression is: ； in, It is The weight of a specific modality data, is the modality-specific data index, is the number of specific modes, It is The historical contribution of a specific modality data, is the contribution of traversing all modal data, is the historical contribution of the 𝑏th specific modal data; The data of each modality are weighted and fused to obtain multimodal comprehensive data, which is expressed as: ； in, It's time Multimodal comprehensive data, It's time No. The actual value of the modal data, is the modal data index, is the number of modal data; Integrate variational autoencoders and convolutional neural networks to capture the global and local features of each modality data; Use variational autoencoders to learn historical normal data, generate potential distributions for each modal data, encode real-time input modal data based on variational autoencoders, and compare them with the potential distributions of historical normal data to calculate the degree of deviation; Use convolutional neural networks to extract local features from modal data and detect whether there are local abnormal signals; The anomaly score of the current multimodal comprehensive data is calculated through the anomaly scoring mechanism based on KL divergence. The expression is: ； in, It is at the moment The abnormal score of It's time Multimodal comprehensive data, It's time The current distribution of multimodal synthetic data, It is The normal distribution of the composite data calculated from the modal data is ) is the moment The current distribution of health prediction values, It's time The health prediction value of It is Normal distribution of health prediction values ​​for modal data, is the time factor, is an asymmetric measure of the difference between two probability distributions.

2. The obstetric real-time nursing optimization method based on big data according to claim 1, characterized in that: Classifying abnormal events based on risk assessment results and generating personalized care plans include the following steps: Abnormal events are classified into low risk, medium risk and high risk according to the size of the abnormality score; When low risk is detected, no immediate intervention is required and personalized advice is provided to pregnant women; When a moderate risk is detected, a notification is automatically sent to the doctor to provide the current abnormal details, and the doctor is advised to observe and further monitor relevant physiological data through the remote monitoring platform, provide nursing guidance, push abnormality reports to pregnant women, and recommend regular clinical examinations; When a high-risk abnormality is detected, the doctor is notified immediately and the pregnant woman is advised to go to the hospital immediately. The frequency of fetal heart monitoring is automatically increased, the physiological signals of the fetus and the mother are tracked in real time, an alarm is sent through the emergency notification platform, and an emergency care plan is initiated.

3. The obstetric real-time nursing optimization method based on big data according to claim 2, characterized in that: Automatic optimization of the generated personalized care plan includes the following steps: Regularly collect new multimodal data and user feedback; Evaluate the effectiveness of current care programs based on new data and user feedback; Based on the evaluation results, personalized suggestions are added and the user experience is improved to form an optimized care plan.

4. 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 obstetric real-time care optimization method based on big data described in any one of claims 1 to 3 are implemented.

5. 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 obstetric real-time care optimization method based on big data described in any one of claims 1 to 3 are implemented.

Images