Pregnant and lying-in woman health management platform based on multi-modal data fusion

Abstract

The invention discloses a pregnant and lying-in woman health management platform based on multi-modal data fusion, and relates to the technical field of monitoring analysis, and the platform comprises a data collection module which is used for collecting physiological index data, living habit data and psychological state data corresponding to a target pregnant woman; the data fusion module is used for fusing the preprocessed physiological index data, the preprocessed living habit data and the preprocessed psychological state data, extracting features corresponding to the physiological index data, the living habit data and the psychological state data, and generating health feature vectors; the health assessment module is used for assessing the health condition of the target pregnant woman according to the health feature vector, and determining a comprehensive health assessment coefficient corresponding to the target pregnant woman; and the early warning module is used for monitoring the comprehensive health assessment coefficient corresponding to the target pregnant woman, outputting early warning information when the comprehensive health assessment coefficient corresponding to the target pregnant woman is monitored to be abnormal, and performing management based on preset adjustment measures. The method has the effect of improving the health management efficiency.

Description

Technical Field

[0001] The present application relates to the field of monitoring and analysis technology, and in particular to a maternal and infant health management platform based on multimodal data fusion. Background Art

[0002] With China's rapid socioeconomic development and profound demographic shifts, declining fertility rates and the fading demographic dividend have become significant challenges to current social development. Furthermore, inadequate postpartum care services and the increasing prevalence of postpartum depression have further increased the burden on families and society. In this context, leveraging artificial intelligence and big data technologies to advance the intelligent transformation of postpartum care and depression monitoring is not only a crucial response to national policy but also a necessary step to meet social needs and improve the quality of healthcare services.

[0003] In related technologies, traditional health management methods mainly rely on single physiological indicator tests, such as blood pressure, blood sugar, etc., which cannot fully reflect the health status of pregnant women, and do not collect and analyze the living habits and psychological state of pregnant women, resulting in a lack of effective data fusion and analysis methods, which effectively reduces management efficiency and leaves room for improvement. Summary of the Invention

[0004] In response to the shortcomings of existing technologies, this application provides a maternal and infant health management platform based on multimodal data fusion.

[0005] First, this application provides a maternal health management platform based on multimodal data fusion, including: A data collection module is used to collect physiological indicator data, living habit data, and psychological state data corresponding to the target pregnant woman; The data processing module is connected to the data acquisition module and is used to pre-process the physiological indicator data, living habit data and psychological state data corresponding to the target pregnant woman; The data fusion module is connected to the data processing module and is used to fuse the pre-processed physiological indicator data, living habit data and psychological state data, and extract the features corresponding to the physiological indicator data, living habit data and psychological state data, and then generate a health feature vector; a health assessment module, connected to the data fusion module, for assessing the health status of the target pregnant woman according to the health feature vector, and determining a comprehensive health assessment coefficient corresponding to the target pregnant woman based on the assessment result; The early warning module is connected to the health assessment module and is used to monitor the comprehensive health assessment coefficient corresponding to the target pregnant woman. When the comprehensive health assessment coefficient corresponding to the target pregnant woman is detected to be abnormal, it outputs early warning information and manages based on preset adjustment measures.

[0006] Preferably, the preprocessing includes data cleaning, data denoising and data normalization.

[0007] Preferably, the pre-processed physiological indicator data, living habit data and psychological state data are fused, and features corresponding to the physiological indicator data, living habit data and psychological state data are extracted to generate a health feature vector, specifically including: Extract features from the target pregnant woman's physiological indicator data, lifestyle data, and psychological state data according to a preset method, and then determine the target user's physiological indicator feature vector, lifestyle feature vector, and psychological state feature vector; The physiological indicator feature vector, life habit feature vector and psychological state feature vector corresponding to the target user are integrated, and the physiological indicator feature vector, life habit feature vector and psychological state feature vector corresponding to the target user are connected in series to form a comprehensive feature vector; The fused comprehensive feature vector is input into the preset deep neural network to generate a health feature vector.

[0008] Preferably, the health status of the target pregnant woman is evaluated according to the health feature vector, and a comprehensive health assessment coefficient corresponding to the target pregnant woman is determined based on the evaluation result, specifically including: Extract physiological index data from the health feature vector , and obtain the preset normal threshold range of physiological indicators ,in, Indicates the number corresponding to each physiological indicator, ; By formula , confirm the physiological health risk coefficient corresponding to the target pregnant woman ; The living habit feature vector is extracted from the health feature vector and quantified as ,in, They are respectively expressed as the quantitative scores of diet, exercise, and sleep corresponding to the target pregnant woman; By formula , confirm the life health coefficient corresponding to the target pregnant woman ,in, Respectively expressed as the scores of preset standard diet, standard exercise, and standard sleep; Extract the psychological state feature vector from the health feature vector, and obtain the score vector after the psychological state feature vector is quantized , and confirm the average score vector corresponding to the psychological state data ; By formula , confirm the mental health coefficient corresponding to the target pregnant woman ,in, It is represented by the number corresponding to each psychological state, ,and ; The physiological health risk coefficient corresponding to the target pregnant woman , the life health coefficient corresponding to the target pregnant woman The mental health coefficient corresponding to the target pregnant woman Substitute into the formula Then confirm the comprehensive health assessment coefficient corresponding to the target pregnant woman ,in, They are respectively represented as preset weight coefficients.

[0009] Preferably, the comprehensive health assessment coefficient corresponding to the target pregnant woman is monitored, and when the comprehensive health assessment coefficient corresponding to the target pregnant woman is detected to be abnormal, an early warning message is output, and management is performed based on preset adjustment measures, specifically including: The comprehensive health assessment coefficient corresponding to the target pregnant woman and the preset comprehensive health assessment threshold range Compare, among which, Represented as the preset first comprehensive health assessment threshold, It is represented as a preset second comprehensive health assessment threshold; If the comprehensive health assessment coefficient of the target pregnant woman , then the target pregnant woman is judged to have no abnormality; If the comprehensive health assessment coefficient of the target pregnant woman in If the level is between 0 and 1, further monitoring is required and a first-level warning signal is output; If the comprehensive health assessment coefficient of the target pregnant woman , the target pregnant woman is judged to be abnormal and a second-level warning signal is output, wherein the warning signal of the second-level warning signal is greater than the first-level warning signal.

[0010] Preferably, the further monitoring process specifically includes: In the preset time period, the comprehensive health assessment coefficients corresponding to the target pregnant women are collected in real time to form a time series, and the comprehensive health assessment coefficients corresponding to the target pregnant women are calculated according to the time series using the function express; By formula , confirm the change coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman ,in, Indicates a preset time period; The change coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman The preset change threshold Make a comparison; If the change coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman , then the target pregnant woman is judged to have no abnormality; If the change coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman , it is determined that the target pregnant woman will have abnormalities in the future and a third-level warning signal is output.

[0011] Preferably, after determining that the target pregnant woman is abnormal and outputting a secondary warning signal, the method further includes: Obtain the comprehensive health assessment coefficient corresponding to the target pregnant woman, and confirm the deviation coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman based on the comprehensive health assessment coefficient of the target pregnant woman, and then confirm the management level corresponding to the comprehensive health assessment coefficient of the target pregnant woman based on the deviation coefficient, and then perform health management on the target pregnant woman based on the management level.

[0012] Preferably, it also includes: Obtain the physiological health risk coefficient corresponding to the target pregnant woman within the preset time period , the life health coefficient corresponding to the target pregnant woman The mental health coefficient corresponding to the target pregnant woman , calculate the preset weight coefficient :

[0013] in, It is expressed as the mean value of the physiological health risk coefficient of the target pregnant woman within the preset time period. It is expressed as the mean value of the target pregnant woman’s life health coefficient within the preset time period. It is expressed as the mean value of the target pregnant woman's mental health coefficient within the preset time period.

[0014] In a second aspect, the present application provides a maternal health management method based on multimodal data fusion, comprising the following steps: Collect physiological indicator data, living habit data and psychological state data corresponding to the target pregnant woman; Pre-processing of physiological indicator data, living habit data, and psychological state data corresponding to the target pregnant woman; Fusing the pre-processed physiological indicator data, lifestyle data, and psychological state data, and extracting the corresponding features of the physiological indicator data, lifestyle data, and psychological state data to generate a health feature vector; Assessing the health status of the target pregnant woman according to the health feature vector, and determining a comprehensive health assessment coefficient corresponding to the target pregnant woman based on the assessment result; The comprehensive health assessment coefficient corresponding to the target pregnant woman is monitored, and when the comprehensive health assessment coefficient corresponding to the target pregnant woman is detected to be abnormal, an early warning message is output, and management is carried out based on preset adjustment measures.

[0015] In a third aspect, the present application provides a computer-readable storage medium storing instructions. When the instructions are executed on a computer, the computer executes any one of the above-mentioned maternal and infant health management platforms based on multimodal data fusion.

[0016] In summary, this application includes at least one of the following beneficial technical effects: 1. This application provides a maternal health management platform based on multimodal data fusion. By collecting and analyzing the physiological indicator data, lifestyle data, and psychological state data corresponding to the target pregnant woman, the platform determines the comprehensive health assessment coefficient corresponding to the target pregnant woman and monitors the comprehensive health assessment coefficient of the target pregnant woman. When the comprehensive health assessment coefficient of the target pregnant woman is detected to be abnormal, an early warning message is output and management is carried out based on preset adjustment measures, thereby effectively improving management efficiency. 2. By collecting the comprehensive health assessment coefficient corresponding to the target pregnant woman in real time, the change coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman is confirmed, and the change coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman is compared with the preset change threshold. Based on the comparison results, early warning management of the target pregnant woman's physical condition in the future period is carried out, and the physical condition of the target pregnant woman is effectively predicted, thereby effectively improving management efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0018] Figure 1 This is a schematic diagram of a maternal and infant health management platform based on multimodal data fusion in an embodiment of the present application.

[0019] Figure 2 This is a flow chart of the method of the maternal and infant health management platform based on multimodal data fusion in an embodiment of the present application. DETAILED DESCRIPTION

[0020] The following is combined with Figure 1-2 This application is described in further detail.

[0021] Example 1 The embodiments of the present application disclose a maternal and infant health management platform based on multimodal data fusion.

[0022] Reference Figure 1 , a maternal and infant health management platform based on multimodal data fusion, including: A data collection module is used to collect physiological indicator data, living habit data, and psychological state data corresponding to the target pregnant woman; The data processing module is connected to the data acquisition module and is used to pre-process the physiological indicator data, living habit data and psychological state data corresponding to the target pregnant woman; The data fusion module is connected to the data processing module and is used to fuse the pre-processed physiological indicator data, living habit data and psychological state data, and extract the features corresponding to the physiological indicator data, living habit data and psychological state data, and then generate a health feature vector; a health assessment module, connected to the data fusion module, for assessing the health status of the target pregnant woman according to the health feature vector, and determining a comprehensive health assessment coefficient corresponding to the target pregnant woman based on the assessment result; The early warning module is connected to the health assessment module and is used to monitor the comprehensive health assessment coefficient corresponding to the target pregnant woman. When the comprehensive health assessment coefficient corresponding to the target pregnant woman is detected to be abnormal, it outputs early warning information and manages based on preset adjustment measures.

[0023] Furthermore, the preprocessing includes data cleaning, data denoising and data normalization.

[0024] Specifically, data cleaning involves checking the collected data for missing values, outliers, and duplicate values. For missing values, if the missing proportion is small, use the mean, median, or mode filling method to fill in the missing values. If the missing proportion is large, select an appropriate method based on the data characteristics and business logic, such as deleting the corresponding data records or using interpolation to estimate the missing values. For outliers, use statistical analysis methods to identify them and handle them based on the specific situation, such as correcting the outliers or deleting abnormal data records. For duplicate values, directly delete the duplicate data records to ensure data uniqueness.

[0025] Data denoising: Filtering algorithms are used to remove noise that may exist in physiological indicator data and lifestyle data. For example, for physiological indicator data, Gaussian filtering and median filtering are used to remove high-frequency noise and make the data smoother. For lifestyle data, sliding average filtering and other methods are used to reduce data fluctuations and improve data stability.

[0026] Data normalization: Normalize data of different types and ranges to make them have the same scale and range.

[0027] The pre-processed physiological indicator data, lifestyle data, and psychological state data are integrated, and the corresponding features of the physiological indicator data, lifestyle data, and psychological state data are extracted to generate a health feature vector, which specifically includes: Extract features from the target pregnant woman's physiological indicator data, lifestyle data, and psychological state data according to a preset method, and then determine the target user's physiological indicator feature vector, lifestyle feature vector, and psychological state feature vector; The physiological indicator feature vector, life habit feature vector and psychological state feature vector corresponding to the target user are integrated, and the physiological indicator feature vector, life habit feature vector and psychological state feature vector corresponding to the target user are connected in series to form a comprehensive feature vector; The fused comprehensive feature vector is input into the preset deep neural network to generate a health feature vector.

[0028] Specifically, a convolutional neural network is used for feature extraction, and the physiological indicator data within a preset time period is used as input data. The physiological indicator feature vector is extracted through the convolution layer, pooling layer and fully connected layer; The living habit data corresponding to the target pregnant woman is extracted using a recurrent neural network, and the living habit data corresponding to the target pregnant woman is input in chronological order to obtain a living habit feature vector; For the psychological state data, a multi-layer perceptron is used to extract features. After the psychological state data is one-hot encoded, it is input into the multi-layer perceptron. Through the calculation of multi-layer neurons and the action of activation functions, the psychological state feature vector is extracted.

[0029] Specifically, the health feature vector integrates the characteristics of multimodal data such as physiological indicators, living habits and psychological state, and can comprehensively and accurately reflect the health status of pregnant women. Through the analysis of this vector, the health assessment module can calculate specific indicators such as physiological health risk coefficient, living habit health index, and mental health level coefficient based on the rich information contained therein and according to the corresponding assessment algorithms and models, and then derive a comprehensive health assessment coefficient to generate a detailed and accurate health assessment report for pregnant women. Compared with single modality data, it greatly improves the accuracy and reliability of health assessment.

[0030] It should be noted that the health status of the target pregnant woman is evaluated based on the health feature vector, and the comprehensive health assessment coefficient corresponding to the target pregnant woman is determined based on the evaluation result, specifically including: Extract physiological index data from the health feature vector , and obtain the preset normal threshold range of physiological indicators ,in, Indicates the number corresponding to each physiological indicator, ; By formula , confirm the physiological health risk coefficient corresponding to the target pregnant woman ; In the embodiment of the present application, physiological indicator data include but are not limited to blood pressure, blood sugar, weight, heart rate, and fetal movement; That is, when the physiological indicators are within the preset normal threshold range, the contribution to the risk coefficient is 0. Conversely, the corresponding risk value is calculated according to the degree of deviation from the center of the normal range. Finally, the risk values of all physiological indicators are averaged to obtain the physiological health risk coefficient. The higher the coefficient, the greater the physiological health risk. The living habit feature vector is extracted from the health feature vector and quantified as ,in, They are respectively expressed as the quantitative scores of diet, exercise, and sleep corresponding to the target pregnant woman; By formula , confirm the life health coefficient corresponding to the target pregnant woman ,in, Respectively expressed as the scores of preset standard diet, standard exercise, and standard sleep; In the embodiment of the present application, living habits include eating habit data, exercise habit data, and sleeping habit data, wherein a higher health coefficient indicates a healthier living habit; Extract the psychological state feature vector from the health feature vector, and obtain the score vector after the psychological state feature vector is quantized , and confirm the average score vector corresponding to the psychological state data ; By formula , confirm the mental health coefficient corresponding to the target pregnant woman ,in, It is represented by the number corresponding to each psychological state, ,and ; The physiological health risk coefficient corresponding to the target pregnant woman , the life health coefficient corresponding to the target pregnant woman The mental health coefficient corresponding to the target pregnant woman Substitute into the formula Then confirm the comprehensive health assessment coefficient corresponding to the target pregnant woman ,in, They are respectively represented as preset weight coefficients.

[0031] Furthermore, the comprehensive health assessment coefficient corresponding to the target pregnant woman is monitored, and when the comprehensive health assessment coefficient corresponding to the target pregnant woman is detected to be abnormal, an early warning message is output, and management is carried out based on preset adjustment measures, specifically including: The comprehensive health assessment coefficient corresponding to the target pregnant woman and the preset comprehensive health assessment threshold range Compare, among which, Represented as the preset first comprehensive health assessment threshold, It is represented as a preset second comprehensive health assessment threshold; If the comprehensive health assessment coefficient of the target pregnant woman , then the target pregnant woman is judged to have no abnormality; If the comprehensive health assessment coefficient of the target pregnant woman in If the level is between 0 and 1, further monitoring is required and a first-level warning signal is output; If the comprehensive health assessment coefficient of the target pregnant woman , the target pregnant woman is judged to be abnormal and a second-level warning signal is output, wherein the warning signal of the second-level warning signal is greater than the first-level warning signal.

[0032] Specifically, the preset adjustment measures include but are not limited to diet, exercise and psychological adjustment. Among them, dietary recommendations are formulated according to the nutritional needs, physical condition and personal tastes of pregnant women, and exercise plans are designed based on the pregnant women's gestational age, physical condition and exercise preferences. Psychological adjustment measures include psychological counseling, relaxation training and psychological counseling course recommendations.

[0033] It should be noted that the further monitoring process specifically includes: In the preset time period, the comprehensive health assessment coefficients corresponding to the target pregnant women are collected in real time to form a time series, and the comprehensive health assessment coefficients corresponding to the target pregnant women are calculated according to the time series using the function express; By formula , confirm the change coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman ,in, Indicates a preset time period; The change coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman The preset change threshold Make a comparison; If the change coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman , then the target pregnant woman is judged to have no abnormality; If the change coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman , it is determined that the target pregnant woman will have abnormalities in the future and a third-level warning signal is output.

[0034] Furthermore, after determining that the target pregnant woman is abnormal and outputting a secondary warning signal, the following steps are also included: Obtain the comprehensive health assessment coefficient corresponding to the target pregnant woman, and confirm the deviation coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman based on the comprehensive health assessment coefficient of the target pregnant woman, and then confirm the management level corresponding to the comprehensive health assessment coefficient of the target pregnant woman based on the deviation coefficient, and then perform health management on the target pregnant woman based on the management level.

[0035] Specifically, the calculation process of the deviation coefficient corresponding to the target pregnant woman's comprehensive health assessment coefficient is as follows: ,in, Expressed as the comprehensive health assessment coefficient corresponding to the target pregnant woman, It is represented as a preset first comprehensive health assessment threshold; Compare the deviation coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman with the preset risk threshold interval; If the deviation coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman is lower than the low-risk threshold, it is determined to be a mild risk level, and the target pregnant woman will undergo routine prenatal examinations; If the deviation coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman is between the low-risk threshold and the high-risk threshold, it is determined to be a moderate risk level, and the frequency of prenatal examinations for the target pregnant woman will be increased, and fetal heart monitoring will be performed; If the deviation coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman is higher than the high-risk threshold, it is judged to be a high-risk level, and the target pregnant woman needs to be managed using an emergency plan and hospitalized for observation.

[0036] Furthermore, it also includes: Obtain the physiological health risk coefficient corresponding to the target pregnant woman within the preset time period , the life health coefficient corresponding to the target pregnant woman The mental health coefficient corresponding to the target pregnant woman , calculate the preset weight coefficient :

[0037] in, It is expressed as the mean value of the physiological health risk coefficient of the target pregnant woman within the preset time period. It is expressed as the mean value of the target pregnant woman’s life health coefficient within the preset time period. It is expressed as the mean value of the target pregnant woman's mental health coefficient within the preset time period.

[0038] Specifically, obtaining the preset weight coefficient can accurately reflect the comprehensive health assessment coefficient corresponding to the target pregnant woman, and then timely discover potential safety hazards based on the comprehensive health assessment coefficient corresponding to the target pregnant woman.

[0039] Example 2 The embodiments of the present application also disclose a maternal and child health management method based on multimodal data fusion.

[0040] Reference Figure 2 , a maternal health management method based on multimodal data fusion, including the following steps: Collect physiological indicator data, living habit data and psychological state data corresponding to the target pregnant woman; Pre-processing of physiological indicator data, living habit data, and psychological state data corresponding to the target pregnant woman; Fusing the pre-processed physiological indicator data, lifestyle data, and psychological state data, and extracting the corresponding features of the physiological indicator data, lifestyle data, and psychological state data to generate a health feature vector; Assessing the health status of the target pregnant woman according to the health feature vector, and determining a comprehensive health assessment coefficient corresponding to the target pregnant woman based on the assessment result; The comprehensive health assessment coefficient corresponding to the target pregnant woman is monitored, and when the comprehensive health assessment coefficient corresponding to the target pregnant woman is detected to be abnormal, an early warning message is output, and management is carried out based on preset adjustment measures.

[0041] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the described specific embodiments or replace them in similar ways. As long as they do not deviate from the concept of the invention, they should all fall within the scope of protection of the present invention.

[0042] Throughout this specification, references to terms such as "one embodiment," "example," or "specific example" indicate that the specific features, structures, materials, or characteristics described in conjunction with that embodiment or example are included in at least one embodiment or example of the present invention. In this specification, schematic representations of these terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0043] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the present invention to specific embodiments. Obviously, many modifications and variations are possible based on the contents of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention.

Claims

1. A maternal and infant health management platform based on multimodal data fusion, characterized by: include: A data collection module is used to collect physiological indicator data, living habit data, and psychological state data corresponding to the target pregnant woman; The data processing module is connected to the data acquisition module and is used to pre-process the physiological indicator data, living habit data and psychological state data corresponding to the target pregnant woman; The data fusion module is connected to the data processing module and is used to fuse the pre-processed physiological indicator data, living habit data and psychological state data, and extract the features corresponding to the physiological indicator data, living habit data and psychological state data, and then generate a health feature vector; a health assessment module, connected to the data fusion module, for assessing the health status of the target pregnant woman according to the health feature vector, and determining a comprehensive health assessment coefficient corresponding to the target pregnant woman based on the assessment result; The early warning module is connected to the health assessment module and is used to monitor the comprehensive health assessment coefficient corresponding to the target pregnant woman. When the comprehensive health assessment coefficient corresponding to the target pregnant woman is detected to be abnormal, it outputs early warning information and manages based on preset adjustment measures.

2. The maternal and infant health management platform based on multimodal data fusion according to claim 1 is characterized in that: The preprocessing includes data cleaning, data denoising and data normalization.

3. The maternal and infant health management platform based on multimodal data fusion according to claim 1 is characterized in that: The pre-processed physiological indicator data, lifestyle data, and psychological state data are integrated, and the corresponding features of the physiological indicator data, lifestyle data, and psychological state data are extracted to generate a health feature vector, which specifically includes: Extract features from the target pregnant woman's physiological indicator data, lifestyle data, and psychological state data according to a preset method, and then determine the target user's physiological indicator feature vector, lifestyle feature vector, and psychological state feature vector; The physiological indicator feature vector, life habit feature vector and psychological state feature vector corresponding to the target user are integrated, and the physiological indicator feature vector, life habit feature vector and psychological state feature vector corresponding to the target user are connected in series to form a comprehensive feature vector; The fused comprehensive feature vector is input into the preset deep neural network to generate a health feature vector.

4. The maternal and infant health management platform based on multimodal data fusion according to claim 3 is characterized in that: The health status of the target pregnant woman is evaluated according to the health feature vector, and a comprehensive health assessment coefficient corresponding to the target pregnant woman is determined based on the evaluation result, specifically including: Extract physiological index data from the health feature vector , and obtain the preset normal threshold range of physiological indicators ,in, Indicates the number corresponding to each physiological indicator, ; By formula , confirm the physiological health risk coefficient corresponding to the target pregnant woman ; The living habit feature vector is extracted from the health feature vector and quantified as ,in, They are respectively expressed as the quantitative scores of diet, exercise, and sleep corresponding to the target pregnant woman; By formula , confirm the life health coefficient corresponding to the target pregnant woman ,in, Respectively expressed as the scores of preset standard diet, standard exercise, and standard sleep; Extract the psychological state feature vector from the health feature vector, and obtain the score vector after the psychological state feature vector is quantized , and confirm the average score vector corresponding to the psychological state data ; By formula , confirm the mental health coefficient corresponding to the target pregnant woman ,in, It is represented by the number corresponding to each psychological state, ,and ; The physiological health risk coefficient corresponding to the target pregnant woman , the life health coefficient corresponding to the target pregnant woman The mental health coefficient corresponding to the target pregnant woman Substitute into the formula Then confirm the comprehensive health assessment coefficient corresponding to the target pregnant woman ,in, They are respectively represented as preset weight coefficients.

5. The maternal and infant health management platform based on multimodal data fusion according to claim 4 is characterized in that: Monitor the comprehensive health assessment coefficient of the target pregnant woman, and when the comprehensive health assessment coefficient of the target pregnant woman is detected to be abnormal, output warning information and manage based on preset adjustment measures, including: The comprehensive health assessment coefficient corresponding to the target pregnant woman and the preset comprehensive health assessment threshold range Compare, among which, Represented as the preset first comprehensive health assessment threshold, It is represented as a preset second comprehensive health assessment threshold; If the comprehensive health assessment coefficient of the target pregnant woman , then the target pregnant woman is judged to have no abnormality; If the comprehensive health assessment coefficient of the target pregnant woman in If the level is between 0 and 1, further monitoring is required and a first-level warning signal is output; If the comprehensive health assessment coefficient of the target pregnant woman , the target pregnant woman is judged to be abnormal and a second-level warning signal is output, wherein the warning signal of the second-level warning signal is greater than the first-level warning signal.

6. The maternal and infant health management platform based on multimodal data fusion according to claim 5 is characterized in that: The further monitoring process includes: In the preset time period, the comprehensive health assessment coefficients corresponding to the target pregnant women are collected in real time to form a time series, and the comprehensive health assessment coefficients corresponding to the target pregnant women are calculated according to the time series using the function express; By formula , confirm the change coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman ,in, Indicates a preset time period; The change coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman The preset change threshold Make a comparison; If the change coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman , then the target pregnant woman is judged to have no abnormality; If the change coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman , it is determined that the target pregnant woman will have abnormalities in the future and a third-level warning signal is output.

7. The maternal and infant health management platform based on multimodal data fusion according to claim 5 is characterized in that: After determining that the target pregnant woman is abnormal and outputting a secondary warning signal, it also includes: Obtain the comprehensive health assessment coefficient corresponding to the target pregnant woman, and confirm the deviation coefficient corresponding to the comprehensive health assessment coefficient of the target pregnant woman based on the comprehensive health assessment coefficient of the target pregnant woman, and then confirm the management level corresponding to the comprehensive health assessment coefficient of the target pregnant woman based on the deviation coefficient, and then perform health management on the target pregnant woman based on the management level.

8. The maternal and infant health management platform based on multimodal data fusion according to claim 4 is characterized in that: Also includes: Obtain the physiological health risk coefficient corresponding to the target pregnant woman within the preset time period , the life health coefficient corresponding to the target pregnant woman The mental health coefficient corresponding to the target pregnant woman , calculate the preset weight coefficient : in, It is expressed as the mean value of the physiological health risk coefficient of the target pregnant woman within the preset time period. It is expressed as the mean value of the target pregnant woman’s life health coefficient within the preset time period. It is expressed as the mean value of the target pregnant woman's mental health coefficient within the preset time period.

9. A maternal health management method based on multimodal data fusion, applied to the maternal health management platform based on multimodal data fusion as described in any one of claims 1 to 8, characterized in that: The following steps are involved: Collect physiological indicator data, living habit data and psychological state data corresponding to the target pregnant woman; Pre-processing of physiological indicator data, living habit data, and psychological state data corresponding to the target pregnant woman; Fusing the pre-processed physiological indicator data, lifestyle data, and psychological state data, and extracting the corresponding features of the physiological indicator data, lifestyle data, and psychological state data to generate a health feature vector; Assessing the health status of the target pregnant woman according to the health feature vector, and determining a comprehensive health assessment coefficient corresponding to the target pregnant woman based on the assessment result; The comprehensive health assessment coefficient corresponding to the target pregnant woman is monitored, and when the comprehensive health assessment coefficient corresponding to the target pregnant woman is detected to be abnormal, an early warning message is output, and management is carried out based on preset adjustment measures.

10. A computer-readable storage medium, characterized in that: Instructions are stored, and when the instructions are run on a computer, the computer is caused to execute the maternal and infant health management platform based on multimodal data fusion as described in any one of claims 1 to 8.

Images