Electronic medical record-based pregnancy result prediction system

Abstract

The invention provides a pregnancy result prediction system based on an electronic medical record. The pregnancy result prediction system based on the electronic medical record comprises a data acquisition terminal, a data transmission device, a cloud terminal and a data prediction terminal, wherein the data acquisition terminal is used for performing real-time acquisition on electronic medical record information of a pregnant user from a database, and calling operation parameters according to data of the database to dynamically adjust and update information query frequency; the data transmission device is used for establishing data communication connection and communication frequency adjustment between the data acquisition terminal and the data prediction terminal, between the data acquisition terminal and the cloud terminal and between the data prediction terminal and the cloud terminal; and the data prediction terminal is used for acquiring a prediction result of pregnancy according to the preprocessed medical record data information and visually displaying the prediction result. And the cloud terminal is used for carrying out cloud storage on the data information sent by the data acquisition terminal and the data prediction terminal.

Description

Technical Field

[0001] The invention proposes a pregnancy outcome prediction system based on electronic medical records, belonging to the technical field of risk assessment based on electronic medical records. Background Art

[0002] With the rapid development of medical information technology, the electronic medical record system has become an indispensable part of the modern medical system. Electronic medical records not only record in detail the patient's diagnosis and treatment process, examination results, medication status and other key information, but also provide rich data resources for clinical decision support, disease prediction and prevention and other fields. In the field of obstetrics and gynecology, the prediction of pregnancy outcomes is of great significance to pregnant women, their families, medical institutions and even the entire society. However, traditional pregnancy outcome prediction methods often rely on doctors' experience judgment or simple statistical analysis, which has problems such as strong subjectivity and insufficient accuracy. In recent years, with the rise of technologies such as big data and artificial intelligence, research on pregnancy outcome prediction based on electronic medical record data has gradually attracted attention. By mining the rich information contained in electronic medical records and combining advanced algorithm models, accurate prediction of pregnancy outcomes can be achieved, providing strong support for clinical decision-making. However, there is currently no complete and efficient pregnancy outcome prediction system based on electronic medical records on the market to meet the high requirements of medical institutions for pregnancy management.

[0003] In the existing technology, although there have been some attempts to use electronic medical record data for disease prediction, most of these studies focus on a single disease or a specific population, and the data processing process is complex and the prediction model is single, which makes it difficult to adapt to the needs of the complex scenario of pregnancy outcome prediction. In addition, the existing systems often lack real-time data collection and dynamic adjustment mechanisms, and cannot reflect changes in the patient's condition in a timely manner, affecting the accuracy and timeliness of the prediction results.

[0004] Therefore, the present invention proposes a pregnancy outcome prediction system based on electronic medical records, which aims to realize real-time collection, efficient transmission and intelligent analysis of electronic medical record information of pregnant users by integrating multiple links such as data collection, transmission, processing and prediction, so as to provide accurate and timely pregnancy outcome prediction services. This system not only solves the problems of insufficient prediction accuracy and low data processing efficiency in the prior art, but also improves the intuitiveness and ease of use of the prediction results through visual display, providing more convenient and efficient pregnancy management services for medical institutions and patients. Summary of the invention

[0005] The present invention provides a pregnancy outcome prediction system based on electronic medical records to solve the technical problems existing in the above-mentioned prior art. The technical solutions adopted are as follows: A pregnancy outcome prediction system based on electronic medical records, comprising a data acquisition terminal, a data transmission device, a cloud terminal and a data prediction terminal; wherein the data acquisition terminal is data-connected to the data prediction terminal via the data transmission device; the data acquisition terminal and the data prediction terminal are data-connected to the cloud terminal via the data transmission device; The data collection terminal is used to collect electronic medical record information of pregnant users from the database in real time, and dynamically adjust the update information query frequency according to the operation parameters retrieved from the database data; The data transmission device is used to establish a data communication connection and adjust the communication frequency between the data collection terminal and the data prediction terminal, between the data collection terminal and the cloud terminal, and between the data prediction terminal and the cloud terminal; The data prediction terminal is used to obtain the pregnancy prediction result according to the pre-processed medical record data information, and to visualize the prediction result; The cloud terminal is used to perform cloud storage on the data information sent by the data collection terminal and the data prediction terminal.

[0006] Furthermore, the data collection terminal comprises: A data update monitoring module, used to monitor in real time whether the electronic medical record information corresponding to the pregnant user in the database has been updated; An update data retrieval module, used to retrieve updated data information when data of the electronic medical record information corresponding to the pregnant user is updated; A data preprocessing module, used to preprocess the updated data information to obtain preprocessed medical record data information; wherein the preprocessing includes data noise reduction processing, data cleaning processing and data standardization processing; The data update retrieval monitoring module is used to monitor the data retrieval operation parameters of the update data retrieval module in real time, and dynamically adjust the update information query frequency of the data update monitoring module on the database according to the data retrieval operation parameters.

[0007] Furthermore, the operation process of the data update and acquisition monitoring module includes: Real-time monitoring of the data retrieval operation parameters of the update data retrieval module, wherein the data retrieval operation parameters include data retrieval completeness ratio, data retrieval response time and data retrieval error occurrence frequency; Obtain data retrieval quality operation index parameters using the data retrieval completeness ratio, data retrieval response time, and data retrieval error occurrence frequency; When the data retrieval quality operation index parameter is lower than a preset operation index parameter threshold, the data update monitoring module adjusts the update information query frequency of the database.

[0008] Furthermore, the data retrieval quality operation index parameters are obtained by using the data retrieval completeness ratio, data retrieval response time and data retrieval error occurrence frequency, including: Compare the complete ratio parameter of the entire data retrieved during the data retrieval process with a preset complete ratio parameter threshold; When the complete ratio parameter of the entire data is lower than a preset complete ratio parameter threshold, the data retrieval response time of the incomplete data is retrieved; The quality operation index parameter is obtained by using the data retrieval response time of the incomplete data combined with the frequency of data retrieval errors; wherein the quality operation index parameter is obtained by the following formula: Among them, S represents the quality operation index parameter; n represents the number of times incomplete data is retrieved; m represents the number of times data is retrieved incorrectly; T c Indicates the preset data retrieval time reference value; T i T represents the data retrieval duration corresponding to the i-th retrieval of incomplete data; j represents the data retrieval time corresponding to the jth data retrieval error; p represents the complete proportion parameter of the entire retrieved data; p 0 Represents the preset full scale parameter threshold.

[0009] Furthermore, when the data retrieval quality operation index parameter is lower than a preset operation index parameter threshold, the data update monitoring module adjusts the update information query frequency of the database, including: When the data retrieval quality operation index parameter is lower than the preset operation index parameter threshold, the update information query frequency of the current data update monitoring module on the database and the minimum value of the data update frequency of the pregnant users in the database are retrieved; Get the concurrent connection number change rate, lock waiting time change rate and lock conflict number change rate corresponding to the current database operation; The database operation load characterization coefficient is obtained by using the change rate of the number of concurrent connections, the change rate of the lock waiting time, and the change rate of the number of lock conflicts corresponding to the current database operation; wherein the database operation load characterization coefficient is obtained by the following formula: Where L represents the database operation load characterization coefficient; S represents the quality operation index parameter; S 0represents the preset threshold of the operating indicator parameter; k represents the number of unit times experienced by the database operation, and the unit time is 1s; N i W represents the change rate of the number of concurrent connections corresponding to the i-th unit time; i represents the rate of change of the lock waiting time corresponding to the i-th unit time; C i Indicates the rate of change of the number of lock conflicts corresponding to the i-th unit time; The adjusted update information query frequency is obtained by using the database operation load characterization coefficient in combination with the update information query frequency of the database and the lowest value of the data update frequency of pregnant users in the database; The adjusted update information query frequency is obtained by the following formula: Among them, F new Indicates the adjusted update information query frequency; F c represents the update information query frequency of the database before adjustment; L represents the database operation load characterization coefficient; F min Indicates the minimum value of the data update frequency of the pregnant users in the database.

[0010] Further, the data transmission device includes a first data transmission device and a second data transmission device; Wherein, the first data transmission device is used to establish a data communication connection between the data acquisition terminal and the data prediction terminal, and dynamically adjust the communication frequency between the data acquisition terminal and the data prediction terminal according to the change of the update information query frequency of the data acquisition terminal; The second data transmission device is used to establish a data communication connection between the data acquisition terminal and the cloud terminal, and between the data prediction terminal and the cloud terminal, and dynamically adjust the communication frequency between the data acquisition terminal and the cloud terminal according to changes in the update information query frequency of the data acquisition terminal.

[0011] Furthermore, the operation process of the first data transmission device includes: Establishing a data communication connection between the data collection terminal and the data prediction terminal; Retrieving the update information query frequency of the data collection terminal in real time, and determining whether the update information query frequency of the data collection terminal has changed; When the update information query frequency of the data acquisition terminal changes, the communication frequency of the communication connection between the data acquisition terminal and the data prediction terminal is adjusted using the changed update information query frequency; The adjusted communication frequency is obtained by the following formula: Among them, Rnew01 Represents the adjusted communication frequency corresponding to the communication connection between the data collection terminal and the data prediction terminal; R c01 Indicates the communication frequency before adjustment corresponding to the communication connection between the data collection terminal and the data prediction terminal; F new Indicates the adjusted update information query frequency; F c represents the update information query frequency of the database before adjustment; L represents the database operation load characterization coefficient; L max It represents the maximum value of the coefficient representing the database operation load that occurs in the database; λ represents the preset attenuation coefficient, which is used to determine the sensitivity of the change in the update information query frequency to the adjustment of the communication frequency; and e represents the base of the natural logarithm.

[0012] Furthermore, the operation process of the second data transmission device includes: Establishing a data communication connection between the data acquisition terminal and the cloud terminal; Establishing a data communication connection between the data prediction terminal and the cloud terminal; Retrieving the update information query frequency of the data collection terminal in real time, and determining whether the update information query frequency of the data collection terminal has changed; When the update information query frequency of the data acquisition terminal changes, the communication frequency of the communication connection between the data acquisition terminal and the data prediction terminal; The communication frequency between the data collection terminal and the cloud terminal is adjusted by using the communication frequency of the communication connection between the data collection terminal and the data prediction terminal in combination with the changed update information query frequency; The adjusted communication frequency corresponding to the communication between the data collection terminal and the cloud terminal is obtained by the following formula: Among them, R new02 Represents the adjusted communication frequency between the data collection terminal and the cloud terminal; R c02 Indicates the communication frequency before adjustment between the data collection terminal and the cloud terminal; F new Represents the adjusted update information query frequency; R new01 represents the adjusted communication frequency corresponding to the communication connection between the data acquisition terminal and the data prediction terminal; L represents the database operation load characterization coefficient; L max It represents the maximum value of the coefficient representing the database operation load that occurs in the database; λ represents the preset attenuation coefficient, which is used to determine the sensitivity of the change in the update information query frequency to the adjustment of the communication frequency; and e represents the base of the natural logarithm.

[0013] Furthermore, the data prediction terminal includes: A real-time receiving monitoring module is used to monitor in real time whether the pre-processed medical record data information is received; A data input module is used to input the pre-processed medical record data information into the deep learning model after training and test optimization when receiving the pre-processed medical record data information; The prediction and visualization display module is used to obtain the prediction results of pregnancy through the deep learning model that has completed training and test optimization in combination with the preprocessed medical record data information, and to visualize the prediction results of pregnancy, wherein the prediction results of pregnancy include whether the pregnancy is likely to proceed smoothly to full term and the prediction of the mode of delivery.

[0014] Furthermore, the structure of the deep learning model is as follows: The input layer is used to receive pre-processed medical record data in real time. These data include but are not limited to the patient's basic information, biochemical indicators, historical medical conditions, examination results, etc.

[0015] Convolutional Neural Networks (CNNs) are used to process image data, such as ultrasound images, to extract spatial features.

[0016] Recurrent Neural Networks (RNNs) are used to process time series data, such as tracking changes in maternal follow-up data.

[0017] Dense Layers: used to process structured data and learn nonlinear relationships between data through multi-layer perceptrons. The feature fusion layer is used to fuse the output data of the convolutional neural network, recurrent neural network and fully connected layer. The key to this layer is to integrate data features from different sources and forms. Simple concatenation or more complex fusion techniques such as feature weighted sum can be used.

[0018] A classification layer, wherein the classification layer includes a binary classification output and a multi-classification output; An output node corresponding to the binary classification output is used to predict whether the pregnancy can proceed smoothly to full term; Another set of output nodes corresponding to the multi-classification output is used to predict the mode of delivery, such as natural delivery, cesarean section, etc. The softmax function can be used to process this multi-classification output; The output layer is used to output specific prediction results, including the probability of full-term pregnancy and the classification results of delivery mode; The post-processing layer is used to interpret the original predictions output by the model and convert them into data information that is easy for doctors and patients to understand.

[0019] Beneficial effects of the present invention: The pregnancy outcome prediction system based on electronic medical records proposed by the present invention realizes comprehensive and real-time collection and intelligent analysis of electronic medical record information of pregnant users by integrating multiple links such as data collection, transmission, processing and prediction, thereby improving the accuracy and reliability of pregnancy outcome prediction. The automated data collection and transmission process reduces manual intervention and errors, and improves the efficiency of data processing and prediction. At the same time, the dynamically adjusted information query frequency and communication frequency further optimize the system performance and reduce resource consumption. The visual display function makes the prediction results more intuitive and easy to understand. Users can quickly understand the pregnancy status and risk prediction without professional knowledge, which enhances user experience and satisfaction. Accurate pregnancy outcome prediction provides doctors with strong decision-making support, enabling them to formulate more personalized and scientific diagnosis and treatment plans, and improve the quality of medical services and patient satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a system block diagram of the pregnancy outcome prediction system based on electronic medical records described in the present invention. DETAILED DESCRIPTION

[0021] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0022] The embodiment of the present invention proposes a pregnancy outcome prediction system based on electronic medical records, such as Figure 1 As shown, the pregnancy outcome prediction system based on electronic medical records includes a data acquisition terminal, a data transmission device, a cloud terminal and a data prediction terminal; wherein the data acquisition terminal is data-connected to the data prediction terminal via the data transmission device; the data acquisition terminal and the data prediction terminal are data-connected to the cloud terminal via the data transmission device; The data collection terminal is used to collect electronic medical record information of pregnant users from the database in real time, and dynamically adjust the update information query frequency according to the operation parameters retrieved from the database data; The data transmission device is used to establish a data communication connection and adjust the communication frequency between the data collection terminal and the data prediction terminal, between the data collection terminal and the cloud terminal, and between the data prediction terminal and the cloud terminal; The data prediction terminal is used to obtain the pregnancy prediction result according to the pre-processed medical record data information, and to visualize the prediction result; The cloud terminal is used to perform cloud storage on the data information sent by the data collection terminal and the data prediction terminal.

[0023] The working principle of the above technical solution is as follows: the terminal first establishes a connection with a medical institution or a related database, and collects the electronic medical record information of the pregnant user in real time through a preset interface or protocol. This information includes but is not limited to the basic information of the pregnant woman (such as age, weight, number of pregnancies, etc.), past medical history, current pregnancy status (such as gestational age, prenatal examination results, etc.), medication records, etc. The data collection terminal also has intelligent analysis capabilities, and can dynamically adjust its information query frequency according to the data retrieval operation parameters in the database (such as data update frequency, query load, etc.) to ensure that the latest data can be obtained in a timely manner while avoiding excessive access pressure on the database.

[0024] Data transmission device: As a bridge for data transmission, this device is responsible for establishing a stable and efficient data communication connection between the data collection terminal, the data prediction terminal and the cloud terminal. It supports a variety of communication protocols and encryption technologies to ensure the security, reliability and real-time performance of data transmission. At the same time, the transmission device also has the function of adjusting the communication frequency. According to the actual needs of data transmission and the network conditions, it dynamically adjusts the data transmission rate and frequency to optimize the data transmission efficiency.

[0025] Cloud terminal: As the center of data storage and processing, the cloud terminal receives data from the data collection terminal and the data prediction terminal. It has powerful data processing capabilities and storage space, and can support the rapid retrieval, analysis and mining of massive electronic medical record information. The cloud terminal is also responsible for data backup and recovery to ensure data security and availability.

[0026] Data prediction terminal: After receiving the pre-processed medical record data information, the terminal uses advanced algorithm models (such as machine learning, deep learning, etc.) to intelligently analyze the data to obtain the predicted results of pregnancy. The predicted results may include the probability of successful pregnancy, the risk of complications, and the recommendation of delivery methods. The data prediction terminal also has a visual display function, which intuitively presents complex prediction results to users in the form of charts, reports, etc., which is convenient for users to understand and make decisions.

[0027] The effect of the above technical solution is: by integrating multiple links such as data collection, transmission, processing and prediction, comprehensive and real-time collection and intelligent analysis of electronic medical record information of pregnant users can be achieved, thereby improving the accuracy and reliability of pregnancy outcome prediction. The automated data collection and transmission process reduces manual intervention and errors, and improves the efficiency of data processing and prediction. At the same time, the dynamically adjusted information query frequency and communication frequency further optimize system performance and reduce resource consumption. The visual display function makes the prediction results more intuitive and easy to understand. Users can quickly understand the pregnancy status and risk prediction without professional knowledge, which enhances user experience and satisfaction. Accurate pregnancy outcome prediction provides doctors with strong decision-making support, enabling them to formulate more personalized and scientific diagnosis and treatment plans, and improve the quality of medical services and patient satisfaction.

[0028] In one embodiment of the present invention, the data acquisition terminal comprises: A data update monitoring module, used to monitor in real time whether the electronic medical record information corresponding to the pregnant user in the database has been updated; An update data retrieval module, used to retrieve updated data information when data of the electronic medical record information corresponding to the pregnant user is updated; A data preprocessing module, used to preprocess the updated data information to obtain preprocessed medical record data information; wherein the preprocessing includes data noise reduction processing, data cleaning processing and data standardization processing; The data update retrieval monitoring module is used to monitor the data retrieval operation parameters of the update data retrieval module in real time, and dynamically adjust the update information query frequency of the data update monitoring module on the database according to the data retrieval operation parameters.

[0029] The operation process of the data update and acquisition monitoring module includes: S1. Real-time monitoring of the data retrieval operation parameters of the update data retrieval module, wherein the data retrieval operation parameters include data retrieval completeness ratio, data retrieval response time and data retrieval error occurrence frequency; S2. Obtaining data retrieval quality operation index parameters using the data retrieval completeness ratio, data retrieval response time, and data retrieval error occurrence frequency; S3. When the data retrieval quality operation index parameter is lower than a preset operation index parameter threshold, the data update monitoring module adjusts the update information query frequency of the database.

[0030] The working principle of the above technical solution is: the module continuously monitors whether the electronic medical record information corresponding to the pregnant user in the database is updated. When data update is detected, it triggers the subsequent data retrieval process.

[0031] Update data retrieval module: Once the data update monitoring module detects data update, the module will immediately retrieve the updated data information from the database. Such data includes but is not limited to newly generated test results, medication records or other medical information related to pregnancy.

[0032] Data preprocessing module: After the data is successfully retrieved, the data preprocessing module will perform a series of preprocessing operations on the updated data to ensure the quality and consistency of the data. These preprocessing steps include data noise reduction (removing noisy data), data cleaning (correcting or deleting erroneous, incomplete or inconsistent data) and data standardization (converting data into a unified format for subsequent analysis).

[0033] Data update retrieval monitoring module: This module is responsible for monitoring the operation status of the update data retrieval module, including real-time monitoring of data retrieval operation parameters. These parameters include data retrieval completeness ratio (measurement of whether the data is complete), data retrieval response time (measurement of data retrieval speed) and data retrieval error frequency (measurement of the error rate during data retrieval). Through these parameters, the module can evaluate the quality of data retrieval and adjust the frequency of the data update monitoring module's query of the database update information accordingly.

[0034] During the monitoring process, if the data retrieval quality operating index parameter (calculated based on the above three parameters) is lower than the preset operating index parameter threshold, it means that there is a problem with the data retrieval efficiency or quality, which may be caused by excessive database load, network delay or other factors. At this time, the data update retrieval monitoring module will trigger the adjustment mechanism to reduce the query frequency of the data update monitoring module to the database, so as to reduce the system burden and improve the overall operation efficiency.

[0035] The effect of the above technical solution is: through real-time monitoring of data updates and immediate retrieval, the real-time and accuracy of the data is ensured, providing the latest and most accurate data basis for pregnancy outcome prediction.

[0036] Optimize data processing flow: The data preprocessing module effectively improves data quality through noise reduction, cleaning and standardization, providing a reliable data source for subsequent data analysis and prediction. The data update retrieval monitoring module effectively avoids the problem of excessive system load caused by excessive queries by real-time monitoring of data retrieval operation parameters and dynamically adjusts the query frequency according to actual conditions, thereby improving the overall performance and stability of the system. By optimizing the data processing and query process, the user waiting time is reduced, the system response speed is improved, and the user experience is enhanced. High-quality data processing and analysis results provide doctors with more accurate and comprehensive pregnancy outcome prediction information, enabling doctors to formulate more scientific and personalized diagnosis and treatment plans.

[0037] In one embodiment of the present invention, the data retrieval completeness ratio, data retrieval response time and data retrieval error frequency are used to obtain data retrieval quality operation index parameters, including: S201, comparing the complete ratio parameter of the entire data retrieved during the data retrieval process with a preset complete ratio parameter threshold; S202, when the complete ratio parameter of the entire data is lower than a preset complete ratio parameter threshold, retrieving the data retrieval response time of the incomplete data; S203, using the data retrieval response time of the incomplete data combined with the frequency of data retrieval errors to obtain quality operation index parameters; wherein the quality operation index parameters are obtained by the following formula: Among them, S represents the quality operation index parameter; n represents the number of times incomplete data is retrieved; m represents the number of times data is retrieved incorrectly; T c Indicates the preset data retrieval time reference value; T i T represents the data retrieval duration corresponding to the i-th retrieval of incomplete data; j represents the data retrieval time corresponding to the jth data retrieval error; p represents the complete proportion parameter of the entire retrieved data; p 0 Represents the preset full scale parameter threshold.

[0038] The working principle of the above technical solution is as follows: First, the system compares the actual data retrieval complete ratio parameter (that is, the ratio of complete data in the actual retrieved data) with the preset complete ratio parameter threshold. The purpose of this step is to determine whether the integrity of the data retrieval meets the basic requirements. If the overall complete ratio parameter of the data is lower than the preset threshold, it means that there is incomplete data retrieval. At this time, the system will further retrieve the data retrieval response time of these incomplete data, that is, the time it takes for the system to successfully obtain the data (regardless of whether the data is complete). Based on the data retrieval response time of incomplete data and the frequency of data retrieval errors, combined with the preset data retrieval time reference value (Tc), the quality operation index parameter (S) is calculated by a specific formula. This parameter comprehensively considers multiple aspects such as data integrity, retrieval efficiency and error rate, and can fully reflect the quality status of data retrieval.

[0039] The effect of the above technical solution is: through specific formulas and parameters, the above technical solution can achieve quantitative evaluation of data retrieval quality, making data quality monitoring more scientific and objective. Through real-time monitoring and calculation of quality operation index parameters, the system can promptly discover problems in the data retrieval process (such as incomplete data, low retrieval efficiency, high error rate, etc.), and take corresponding measures to solve them, ensuring the quality and availability of data.

[0040] Based on the feedback from the quality operation indicator parameters, the system can dynamically adjust the data retrieval strategy (such as adjusting the query frequency, optimizing the query algorithm, etc.) to reduce the system burden, improve the data retrieval efficiency, and thus optimize the performance of the entire system. High-quality data retrieval services can reduce user waiting time and improve data accuracy, thereby improving user experience and satisfaction. Accurate and reliable data is the basis of intelligent decision-making. Through the above technical solutions, medical institutions can obtain more comprehensive and accurate patient data, providing more powerful support for clinical decision-making.

[0041] In one embodiment of the present invention, when the data retrieval quality operation index parameter is lower than a preset operation index parameter threshold, the data update monitoring module adjusts the update information query frequency of the database, including: S301, when the data retrieval quality operation index parameter is lower than the preset operation index parameter threshold, retrieve the update information query frequency of the current data update monitoring module on the database and the minimum value of the data update frequency of the pregnant users in the database; S302, retrieve the concurrent connection number change rate, lock waiting time change rate and lock conflict number change rate corresponding to the current database operation; S303, using the change rate of the number of concurrent connections, the change rate of the lock waiting time, and the change rate of the number of lock conflicts corresponding to the current database operation to obtain the database operation load characterization coefficient; wherein the database operation load characterization coefficient is obtained by the following formula: Where L represents the database operation load characterization coefficient; S represents the quality operation index parameter; S 0 represents the preset threshold of the operating indicator parameter; k represents the number of unit times experienced by the database operation, and the unit time is 1s; N i W represents the change rate of the number of concurrent connections corresponding to the i-th unit time; i represents the rate of change of the lock waiting time corresponding to the i-th unit time; C i Indicates the rate of change of the number of lock conflicts corresponding to the i-th unit time; S304, obtaining an adjusted update information query frequency by using the database operation load characterization coefficient in combination with the update information query frequency of the database and the lowest value of the data update frequency of pregnant users in the database; The adjusted update information query frequency is obtained by the following formula: Among them, F new Indicates the adjusted update information query frequency; F c represents the update information query frequency of the database before adjustment; L represents the database operation load characterization coefficient; F min Indicates the minimum value of the data update frequency of the pregnant users in the database.

[0042] The working principle of the above technical solution is: when the data retrieval quality operating indicator parameter is lower than the preset operating indicator parameter threshold, the system considers that the current data retrieval quality does not meet the requirements and needs to adjust the query frequency of the data update monitoring module.

[0043] The system first retrieves the update information query frequency of the current data update monitoring module (i.e. the query frequency before adjustment) and the minimum data update frequency of pregnant users in the database. This minimum value reflects the slowest data update speed of pregnant users in the database and is an important factor to be considered when adjusting the query frequency.

[0044] The system evaluates the database operation load by calling key indicators such as the change rate of concurrent connections, lock waiting time, and lock conflict times corresponding to the current database operation. These indicators can reflect the performance bottlenecks and potential problems of the database when processing query requests. Then, these indicators are used to calculate the database operation load characterization coefficient (L) through a specific formula, which quantifies the current operation load status of the database.

[0045] Finally, the system combines the database operation load characterization coefficient (L), the query frequency before adjustment (Fc), and the minimum data update frequency of pregnant users in the database (Fmin), and calculates the adjusted update information query frequency (Fnew) through a specific formula. This adjustment process aims to reduce the database operation load while ensuring the real-time performance of the data, and improve the overall performance and stability of the system.

[0046] The effect of the above technical solution is: by real-time monitoring of data retrieval quality operation index parameters and database operation load, dynamically adjusting the query frequency of the data update monitoring module, both ensuring the real-time and accuracy of the data, and avoiding the degradation of database performance due to excessive query. By reasonably adjusting the query frequency, the burden on the database is reduced, the lock waiting time and the number of lock conflicts are reduced, and the concurrent processing capability and overall performance of the system are improved. At the same time, the system instability caused by query conflicts and data inconsistencies is also reduced. Under the premise of ensuring data quality, the system response speed and user experience are improved by optimizing the query frequency. Users can obtain the required pregnancy user data faster, providing more timely and accurate information support for clinical decision-making. Through the above technical solution, medical institutions can obtain more accurate, timely and comprehensive pregnancy user data, providing more powerful support for clinical decision-making. At the same time, the system can also be intelligently adjusted and optimized according to the actual situation of the data and the operating status of the database, further improving the level of intelligent decision-making.

[0047] In one embodiment of the present invention, the data transmission device includes a first data transmission device and a second data transmission device; Wherein, the first data transmission device is used to establish a data communication connection between the data acquisition terminal and the data prediction terminal, and dynamically adjust the communication frequency between the data acquisition terminal and the data prediction terminal according to the change of the update information query frequency of the data acquisition terminal; The second data transmission device is used to establish a data communication connection between the data acquisition terminal and the cloud terminal, and between the data prediction terminal and the cloud terminal, and dynamically adjust the communication frequency between the data acquisition terminal and the cloud terminal according to changes in the update information query frequency of the data acquisition terminal.

[0048] Specifically, the operation process of the first data transmission device includes: Step a1: establishing a data communication connection between the data acquisition terminal and the data prediction terminal; Step a2: retrieve the update information query frequency of the data acquisition terminal in real time, and determine whether the update information query frequency of the data acquisition terminal changes; Step a3: when the update information query frequency of the data acquisition terminal changes, the communication frequency of the communication connection between the data acquisition terminal and the data prediction terminal is adjusted using the changed update information query frequency; The adjusted communication frequency is obtained by the following formula: Among them, R new01 Represents the adjusted communication frequency corresponding to the communication connection between the data collection terminal and the data prediction terminal; Rc01 Indicates the communication frequency before adjustment corresponding to the communication connection between the data collection terminal and the data prediction terminal; F new Indicates the adjusted update information query frequency; F c represents the update information query frequency of the database before adjustment; L represents the database operation load characterization coefficient; L max It represents the maximum value of the coefficient representing the database operation load that occurs in the database; λ represents the preset attenuation coefficient, which is used to determine the sensitivity of the change in the update information query frequency to the adjustment of the communication frequency; and e represents the base of the natural logarithm.

[0049] The working principle of the above technical solution is: the first data transmission device first establishes a stable data communication connection between the data acquisition terminal and the data prediction terminal to ensure that data can be transmitted smoothly between the two. The device retrieves the update information query frequency of the data acquisition terminal in real time and determines whether this frequency has changed. This step is the basis for dynamically adjusting the communication frequency, because changes in the update information query frequency may reflect the degree of demand for real-time data from the data acquisition terminal. When it is detected that the update information query frequency of the data acquisition terminal has changed, the first data transmission device will use a specific formula to adjust the communication frequency of the communication connection between the data acquisition terminal and the data prediction terminal according to the changed query frequency. This adjustment process is designed to match the communication frequency with the data demand, which not only ensures the real-time transmission of data, but also avoids unnecessary waste of communication resources.

[0050] The adjusted communication frequency is calculated by a formula that comprehensively considers the communication frequency before adjustment, the updated information query frequency after adjustment, the database updated information query frequency before adjustment, the database operation load characterization coefficient and its maximum value, and the preset attenuation coefficient. The design of this formula takes into account multiple factors, so that the adjusted communication frequency can reflect the changes in data demand and take into account the database operation load and system stability.

[0051] The effect of the above technical solution is: when the data retrieval quality operation index parameter is lower than the preset operation index parameter threshold (step S301), the above technical solution can automatically trigger the adjustment mechanism, and dynamically adjust the update information query frequency according to the current database operation status and data update frequency. This adaptive adjustment can ensure that the database can respond and optimize quickly when facing performance bottlenecks or data quality degradation. Comprehensively consider the database operation load: In steps S302 and S303, the above technical solution comprehensively evaluates the database operation load by monitoring the change rate of the number of concurrent connections, the change rate of lock waiting time, and the change rate of the number of lock conflicts of the database. These indicators can reflect the load conditions and potential performance bottlenecks of the database when processing requests. By calculating the database operation load characterization coefficient L, the above technical solution can quantify the current operation status of the database and provide a scientific basis for subsequent adjustments. Optimize the update information query frequency: In step S304, the above technical solution uses the database operation load characterization coefficient L, the database update information query frequency Fc, and the minimum value of the data update frequency of pregnant users in the database F min , calculate the adjusted update information query frequency F through the above formula new This adjustment strategy aims to minimize the use of database resources and improve the overall operation efficiency of the database while ensuring the timeliness of data updates. Protecting special user groups: By considering the minimum data update frequency F of pregnant users in the database min , the above technical solution can ensure that the data update needs of special user groups are met. This is particularly important for medical and health management applications that require real-time monitoring and updating of pregnancy user data, and can improve user satisfaction and trust. Improve system stability and reliability: By dynamically adjusting the query frequency of update information, the above technical solution can reduce the pressure on the database when it is running under high load, and reduce the risk of system crashes or performance degradation caused by untimely data updates or excessive query frequencies. This can improve the stability and reliability of the system and ensure the continuity and availability of data services.

[0052] In addition, by dynamically adjusting the communication frequency, data transmission is matched with data demand in real time, avoiding the waste of communication resources when data demand is low, and improving the overall efficiency of data transmission. Adjusting the communication frequency according to the change of the update information query frequency can reduce unnecessary communication overhead, reduce the system burden, and optimize the utilization of system resources. By considering the database operation load characterization coefficient and the preset attenuation coefficient, the adjustment process is smoother and more stable, reducing system fluctuations and potential problems caused by frequent adjustment of communication frequency. Real-time and accurate data transmission provides reliable data support for data prediction terminals, enabling medical institutions to make more scientific and reasonable clinical decisions. The optimized data transmission service can reduce user waiting time, improve the timeliness and accuracy of data acquisition, and thus enhance the user experience.

[0053] In summary, the above technical solution optimizes and improves database performance by dynamically adjusting the update information query frequency of the data update monitoring module by comprehensively considering the data retrieval quality, database operation load and the needs of special user groups. This can ensure the accuracy and timeliness of data services and improve the overall performance of the system and user experience.

[0054] In one embodiment of the present invention, the operation process of the second data transmission device includes: Step b1, establishing a data communication connection between the data acquisition terminal and the cloud terminal; Step b2: establishing a data communication connection between the data prediction terminal and the cloud terminal; Step b3, retrieving the update information query frequency of the data acquisition terminal in real time, and determining whether the update information query frequency of the data acquisition terminal has changed; Step b4: when the update information query frequency of the data acquisition terminal changes, the communication frequency of the communication connection between the data acquisition terminal and the data prediction terminal; Step b5, adjusting the communication frequency between the data collection terminal and the cloud terminal by using the communication frequency of the communication connection between the data collection terminal and the data prediction terminal combined with the changed update information query frequency; The adjusted communication frequency corresponding to the communication between the data collection terminal and the cloud terminal is obtained by the following formula: Among them, R new02 Represents the adjusted communication frequency between the data collection terminal and the cloud terminal; R c02 Indicates the communication frequency before adjustment between the data collection terminal and the cloud terminal; F new Represents the adjusted update information query frequency; R new01represents the adjusted communication frequency corresponding to the communication connection between the data acquisition terminal and the data prediction terminal; L represents the database operation load characterization coefficient; L max It represents the maximum value of the coefficient representing the database operation load that occurs in the database; λ represents the preset attenuation coefficient, which is used to determine the sensitivity of the change in the update information query frequency to the adjustment of the communication frequency; and e represents the base of the natural logarithm.

[0055] The working principle of the above technical solution is as follows: the second data transmission device first establishes data communication connections between the data acquisition terminal and the cloud terminal, and between the data prediction terminal and the cloud terminal, respectively, to ensure smooth data transmission between these terminals. The device retrieves the update information query frequency of the data acquisition terminal in real time, and determines whether this frequency has changed. This is the basis for subsequent adjustment of the communication frequency between the data acquisition terminal and the cloud terminal.

[0056] When it is detected that the update information query frequency of the data acquisition terminal changes, although the description of step b4 directly mentions adjusting the communication frequency between the data acquisition terminal and the data prediction terminal, according to the context, it is more likely to illustrate that the change in the update information query frequency is a condition that triggers the adjustment of the communication frequency, and the actual adjustment operation will be for the communication frequency between the data acquisition terminal and the cloud terminal. Based on the adjusted communication frequency of the communication connection between the data acquisition terminal and the data prediction terminal, the change in the update information query frequency of the data acquisition terminal, and the database operation load characterization coefficient and other factors, a specific formula is used to calculate and adjust the communication frequency between the data acquisition terminal and the cloud terminal. This adjustment process is intended to enable the cloud terminal to more efficiently receive and process data from the data acquisition terminal, while taking into account the communication needs between the data prediction terminal and the cloud terminal and the database operation load.

[0057] The effect of the above technical solution is: by real-time monitoring of the update information query frequency and adjusting the communication frequency accordingly, data transmission is made more flexible and can quickly adapt to changes in data requirements. The communication frequency is dynamically adjusted according to data requirements to avoid unnecessary waste of communication resources and optimize the allocation and utilization of system resources. By considering the database operation load characterization coefficient and the preset attenuation coefficient, the adjustment process is smoother, reducing system fluctuations and potential problems caused by frequent adjustments to the communication frequency, and enhancing the stability of the system. The adjusted communication frequency can better match data requirements and improve the efficiency and accuracy of data transmission. By optimizing the data transmission mechanism, it provides strong support for cloud terminals to process large amounts of data, enabling medical institutions to better use big data for clinical decision-making and research.

[0058] At the same time, the above technical solution can monitor the update information query frequency of the data acquisition terminal in real time, and dynamically adjust the communication frequency between the data acquisition terminal and the cloud terminal according to its changes. This dynamic adaptability helps to ensure the timeliness and efficiency of data transmission, especially when data updates are frequent or the amount of data is large. By adjusting the communication frequency, the above technical solution can utilize network resources more effectively. When the update information query frequency increases, increasing the communication frequency can ensure timely data transmission; when the update information query frequency decreases, reducing the communication frequency can reduce unnecessary network occupancy, thereby optimizing resource utilization. When adjusting the communication frequency, the above technical solution also takes into account the database operation load characterization coefficient L, which helps to avoid performance degradation caused by increasing the communication frequency when the system is under high load. By introducing L max and the preset attenuation coefficient λ, the above technical solution can more finely control the adjustment range of the communication frequency to ensure the stable operation of the system. By optimizing the communication frequency between the data acquisition terminal and the cloud terminal, the above technical solution can reduce the delay in data transmission and improve the response speed of the system. This is particularly important for application scenarios that require real-time data processing and decision support. In addition, the design of the above technical solution takes into account the scalability of the system. As the number of data acquisition terminals increases or the amount of data grows, the system can maintain efficient operation by dynamically adjusting the communication frequency without the need for large-scale upgrades of hardware or software. By optimizing data transmission efficiency and system performance, the above technical solution can enhance user experience. For example, in application scenarios that require real-time data updates, users can obtain the required information more quickly, thereby improving satisfaction and trust.

[0059] In summary, the above technical solution optimizes data transmission efficiency and system performance by dynamically adjusting the communication frequency between the data collection terminal, the data prediction terminal and the cloud terminal. This adjustment is based on the change of real-time data update frequency and system load considerations, which helps to ensure the timeliness and accuracy of data transmission, while optimizing resource utilization and system response speed.

[0060] In one embodiment of the present invention, the data prediction terminal includes: A real-time receiving monitoring module is used to monitor in real time whether the pre-processed medical record data information is received; A data input module is used to input the pre-processed medical record data information into the deep learning model after training and test optimization when receiving the pre-processed medical record data information; The prediction and visualization module is used to obtain the prediction results of pregnancy through the deep learning model that has been trained and tested and optimized, combined with the pre-processed medical record data information, and visualize the prediction results of pregnancy, wherein the prediction results of pregnancy include whether the pregnancy is likely to go smoothly to full term and the prediction of the mode of delivery. Specific modes of delivery include natural delivery and cesarean section;

[0061] The structure of the deep learning model is as follows: The input layer is used to receive pre-processed medical record data in real time. These data include but are not limited to the patient's basic information, biochemical indicators, historical medical conditions, examination results, etc.

[0062] Convolutional Neural Networks (CNNs) are used to process image data, such as ultrasound images, to extract spatial features.

[0063] Recurrent Neural Networks (RNNs) are used to process time series data, such as tracking changes in maternal follow-up data.

[0064] Dense Layers: used to process structured data and learn nonlinear relationships between data through multi-layer perceptrons. The feature fusion layer is used to fuse the output data of the convolutional neural network, recurrent neural network and fully connected layer. The key to this layer is to integrate data features from different sources and forms. Simple concatenation or more complex fusion techniques such as feature weighted sum can be used.

[0065] A classification layer, wherein the classification layer includes a binary classification output and a multi-classification output; wherein, An output node corresponding to the binary classification output is used to predict whether the pregnancy can proceed to full term; Another set of output nodes corresponding to the multi-classification output is used to predict the mode of delivery, such as natural delivery, cesarean section, etc. The softmax function can be used to process this multi-classification output; The output layer is used to output specific prediction results, including the probability of full-term pregnancy and the classification results of delivery mode; The post-processing layer is used to interpret the original predictions output by the model and convert them into data information that is easy for doctors and patients to understand.

[0066] The working principle of the above technical solution is: the real-time receiving monitoring module continuously monitors whether pre-processed medical record data information has arrived. These data have been preliminarily processed to facilitate the input and analysis of subsequent models.

[0067] Data input: Once the pre-processed medical record data is received, the data input module immediately inputs the data into the deep learning model that has completed training and test optimization. These data include but are not limited to the patient's basic information, biochemical indicators, historical medical conditions, examination results, etc., as well as possible image data (such as ultrasound images) and time series data (such as follow-up data).

[0068] Model processing: After the input layer receives data, the data is distributed to different network layers for processing.

[0069] Convolutional Neural Networks (CNNs) are specialized in processing image data, such as ultrasound images, to extract spatial features from them.

[0070] Recurrent neural network (RNN) processes time series data and captures the temporal variation characteristics in maternal follow-up data.

[0071] Dense Layers process structured data and learn the nonlinear relationship between data through multi-layer perceptrons.

[0072] The feature fusion layer integrates the output data from different network layers to form a comprehensive data feature representation.

[0073] The classification layer includes binary classification output and multi-classification output. The binary classification output predicts whether the pregnancy is likely to go smoothly to full term, while the multi-classification output predicts the mode of delivery (such as natural delivery, cesarean section, etc.), and the softmax function is used to process the multi-classification output.

[0074] Prediction and visualization: The prediction and visualization module receives the output of the classification layer, generates specific prediction results, and displays them to doctors and patients in a visual way. This includes the probability estimation of the possibility of full-term pregnancy and the classification results of the delivery mode.

[0075] Post-processing: The post-processing layer further interprets and transforms the raw predictions output by the model to make them more understandable to doctors and patients. This may include converting probabilities into more intuitive expressions or providing detailed explanations of classification results.

[0076] The effect of the above technical solution is: by integrating multiple deep learning technologies (CNN, RNN, Dense Layers) and feature fusion strategies, it is possible to more comprehensively capture useful information in medical record data, thereby improving the accuracy of pregnancy prediction. Because the model can process multiple types of data (including images, time series, and structured data), it can better support personalized predictions, taking into account the unique situation of each pregnant woman. Real-time reception and rapid processing of medical record data enable doctors to quickly obtain pregnancy prediction results, thereby formulating and adjusting medical plans more quickly. By visually displaying the prediction results, doctors and patients can more intuitively understand the prediction content and enhance communication and trust between doctors and patients. Accurate pregnancy prediction enables medical institutions to better plan and manage resources, ensuring that there are sufficient resources to support the health of pregnant women and newborns during critical periods.

[0077] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A pregnancy outcome prediction system based on electronic medical records, characterized in that: The pregnancy outcome prediction system based on electronic medical records includes a data acquisition terminal, a data transmission device, a cloud terminal and a data prediction terminal; wherein the data acquisition terminal is data-connected to the data prediction terminal via the data transmission device; the data acquisition terminal and the data prediction terminal are data-connected to the cloud terminal via the data transmission device; The data collection terminal is used to collect electronic medical record information of pregnant users from the database in real time, and dynamically adjust the update information query frequency according to the operation parameters retrieved from the database data; The data transmission device is used to establish a data communication connection and adjust the communication frequency between the data acquisition terminal and the data prediction terminal, between the data acquisition terminal and the cloud terminal, and between the data prediction terminal and the cloud terminal; The data prediction terminal is used to obtain the pregnancy prediction result according to the pre-processed medical record data information, and to visualize the prediction result; The cloud terminal is used to perform cloud storage on the data information sent by the data collection terminal and the data prediction terminal.

2. The pregnancy outcome prediction system based on electronic medical records according to claim 1, characterized in that: The data acquisition terminal comprises: A data update monitoring module, used to monitor in real time whether the electronic medical record information corresponding to the pregnant user in the database has been updated; An update data retrieval module, used to retrieve updated data information when data of the electronic medical record information corresponding to the pregnant user is updated; A data preprocessing module, used to preprocess the updated data information to obtain preprocessed medical record data information; wherein the preprocessing includes data noise reduction processing, data cleaning processing and data standardization processing; The data update retrieval monitoring module is used to monitor the data retrieval operation parameters of the update data retrieval module in real time, and dynamically adjust the update information query frequency of the data update monitoring module on the database according to the data retrieval operation parameters.

3. The pregnancy outcome prediction system based on electronic medical records according to claim 2, characterized in that: The operation process of the data update and acquisition monitoring module includes: Real-time monitoring of the data retrieval operation parameters of the update data retrieval module, wherein the data retrieval operation parameters include data retrieval completeness ratio, data retrieval response time and data retrieval error occurrence frequency; Obtain data retrieval quality operation index parameters using the data retrieval completeness ratio, data retrieval response time, and data retrieval error occurrence frequency; When the data retrieval quality operation index parameter is lower than a preset operation index parameter threshold, the data update monitoring module adjusts the update information query frequency of the database.

4. The pregnancy outcome prediction system based on electronic medical records according to claim 3, characterized in that: The data retrieval quality operation index parameters are obtained by using the data retrieval completeness ratio, data retrieval response time and data retrieval error frequency, including: Compare the complete ratio parameter of the entire data retrieved during the data retrieval process with a preset complete ratio parameter threshold; When the complete ratio parameter of the entire data is lower than a preset complete ratio parameter threshold, the data retrieval response time of the incomplete data is retrieved; The quality operation index parameter is obtained by using the data retrieval response time of the incomplete data combined with the frequency of data retrieval errors; wherein the quality operation index parameter is obtained by the following formula: Among them, S represents the quality operation index parameter; n represents the number of times incomplete data is retrieved; m represents the number of times data is retrieved incorrectly; T c Indicates the preset data retrieval time reference value; T i T represents the data retrieval duration corresponding to the i-th retrieval of incomplete data; j represents the data retrieval time corresponding to the j-th data retrieval error; p represents the complete proportion parameter of the entire retrieved data; p0 represents the preset complete proportion parameter threshold.

5. The pregnancy outcome prediction system based on electronic medical records according to claim 3, characterized in that: When the data retrieval quality operation index parameter is lower than a preset operation index parameter threshold, the data update monitoring module adjusts the update information query frequency of the database, including: When the data retrieval quality operation index parameter is lower than the preset operation index parameter threshold, the update information query frequency of the current data update monitoring module on the database and the minimum value of the data update frequency of the pregnant users in the database are retrieved; Get the concurrent connection number change rate, lock waiting time change rate and lock conflict number change rate corresponding to the current database operation; The database operation load characterization coefficient is obtained by using the change rate of the number of concurrent connections, the change rate of the lock waiting time, and the change rate of the number of lock conflicts corresponding to the current database operation; wherein the database operation load characterization coefficient is obtained by the following formula: Wherein, L represents the database operation load characterization coefficient; S represents the quality operation index parameter; S0 represents the preset operation index parameter threshold; k represents the number of unit times experienced by the database operation, and the unit time is 1s; N i W represents the change rate of the number of concurrent connections corresponding to the i-th unit time; i represents the rate of change of the lock waiting time corresponding to the i-th unit time; C i Indicates the rate of change of the number of lock conflicts corresponding to the i-th unit time; The adjusted update information query frequency is obtained by using the database operation load characterization coefficient in combination with the update information query frequency of the database and the lowest value of the data update frequency of pregnant users in the database; The adjusted update information query frequency is obtained by the following formula: Among them, F new Indicates the adjusted update information query frequency; F c represents the update information query frequency of the database before adjustment; L represents the database operation load characterization coefficient; F min Indicates the minimum value of the data update frequency of the pregnant users in the database.

6. The pregnancy outcome prediction system based on electronic medical records according to claim 1, characterized in that: The data transmission device comprises a first data transmission device and a second data transmission device; Wherein, the first data transmission device is used to establish a data communication connection between the data acquisition terminal and the data prediction terminal, and dynamically adjust the communication frequency between the data acquisition terminal and the data prediction terminal according to the change of the update information query frequency of the data acquisition terminal; The second data transmission device is used to establish a data communication connection between the data acquisition terminal and the cloud terminal, and between the data prediction terminal and the cloud terminal, and dynamically adjust the communication frequency between the data acquisition terminal and the cloud terminal according to changes in the update information query frequency of the data acquisition terminal.

7. The pregnancy outcome prediction system based on electronic medical records according to claim 6, characterized in that: The operation process of the first data transmission device includes: Establishing a data communication connection between the data collection terminal and the data prediction terminal; Retrieving the update information query frequency of the data collection terminal in real time, and determining whether the update information query frequency of the data collection terminal has changed; When the update information query frequency of the data acquisition terminal changes, the communication frequency of the communication connection between the data acquisition terminal and the data prediction terminal is adjusted using the changed update information query frequency; The adjusted communication frequency is obtained by the following formula: Among them, R new01 Represents the adjusted communication frequency corresponding to the communication connection between the data collection terminal and the data prediction terminal; R c01 Indicates the communication frequency before adjustment corresponding to the communication connection between the data collection terminal and the data prediction terminal; F new Indicates the adjusted update information query frequency; F c represents the update information query frequency of the database before adjustment; L represents the database operation load characterization coefficient; L max It represents the maximum value of the coefficient representing the database operation load that occurs in the database; λ represents the preset attenuation coefficient, which is used to determine the sensitivity of the change in the update information query frequency to the adjustment of the communication frequency; and e represents the base of the natural logarithm.

8. The pregnancy outcome prediction system based on electronic medical records according to claim 6, characterized in that: The operation process of the second data transmission device includes: Establishing a data communication connection between the data acquisition terminal and the cloud terminal; Establishing a data communication connection between the data prediction terminal and the cloud terminal; Retrieving the update information query frequency of the data collection terminal in real time, and determining whether the update information query frequency of the data collection terminal has changed; When the update information query frequency of the data acquisition terminal changes, the communication frequency of the communication connection between the data acquisition terminal and the data prediction terminal; The communication frequency between the data collection terminal and the cloud terminal is adjusted by using the communication frequency of the communication connection between the data collection terminal and the data prediction terminal in combination with the changed update information query frequency; The adjusted communication frequency corresponding to the communication between the data collection terminal and the cloud terminal is obtained by the following formula: Among them, R new02 Represents the adjusted communication frequency between the data collection terminal and the cloud terminal; R c02 Indicates the communication frequency before adjustment between the data collection terminal and the cloud terminal; F new Represents the adjusted update information query frequency; R new01 represents the adjusted communication frequency corresponding to the communication connection between the data acquisition terminal and the data prediction terminal; L represents the database operation load characterization coefficient; L max It represents the maximum value of the coefficient representing the database operation load that occurs in the database; λ represents the preset attenuation coefficient, which is used to determine the sensitivity of the change in the update information query frequency to the adjustment of the communication frequency; and e represents the base of the natural logarithm.

9. The pregnancy outcome prediction system based on electronic medical records according to claim 1, characterized in that: The data prediction terminal comprises: A real-time receiving monitoring module is used to monitor in real time whether the pre-processed medical record data information is received; A data input module is used to input the pre-processed medical record data information into the deep learning model after training and test optimization when receiving the pre-processed medical record data information; The prediction and visualization display module is used to obtain the prediction results of pregnancy through the deep learning model that has completed training and test optimization in combination with the preprocessed medical record data information, and to visualize the prediction results of pregnancy, wherein the prediction results of pregnancy include whether the pregnancy is likely to proceed smoothly to full term and the prediction of the mode of delivery.

10. The pregnancy outcome prediction system based on electronic medical records according to claim 9, characterized in that: The structure of the deep learning model is as follows: The input layer is used to receive pre-processed medical record data in real time; Convolutional neural networks, which are used to process image data; Recurrent neural networks, for processing time series data; The fully connected layer is used to process structured data and learn the nonlinear relationship between data through a multi-layer perceptron; Feature fusion layer, used to fuse the output data of convolutional neural network, recurrent neural network and fully connected layer; A classification layer, wherein the classification layer includes a binary classification output and a multi-classification output; An output node corresponding to the binary classification output is used to predict whether the pregnancy can proceed smoothly to full term; Another set of output nodes corresponding to the multi-classification output is used to predict the mode of delivery; The output layer is used to output specific prediction results, including the probability of full-term pregnancy and the classification results of delivery mode; The post-processing layer is used to interpret the original predictions output by the model and convert them into data information that is easy for doctors and patients to understand.

Images