Pregnancy risk assessment system for pregnant women based on previous electronic medical history

Abstract

This invention relates to the field of medical data processing technology, specifically to a pregnancy risk assessment system for pregnant women based on their previous electronic medical history. For any important vital sign parameter, this invention obtains the degree of abnormality of the detection state in a local range at each moment; combines the time difference between adjacent moments to obtain an index of the change in detection state between adjacent moments; obtains normal time periods and initial abnormal time periods for temporal feature analysis, and combines the corresponding detection state change index within the initial abnormal time period to obtain an abnormality degree state coefficient for each initial abnormal time period, thus filtering out actual abnormal time periods; analyzes the detection data of each important vital sign parameter in the neighborhood of the current moment or the neighborhood of the next moment to obtain the abnormality degree state coefficient for each actual abnormal time period, thus filtering out highly abnormal time periods. This invention improves the accuracy of pregnancy risk assessment by obtaining important vital sign parameters and accurate distribution of abnormal states.

Description

Technical Field

[0001] This invention relates to the field of medical data processing technology, specifically to a pregnancy risk assessment system for pregnant women based on their previous electronic medical history. Background Technology

[0002] Pregnancy is a crucial stage in a woman's physiological journey. During pregnancy, the mother's physiological functions undergo significant changes, and a series of pregnancy complications may occur, such as hypertension, posing potential risks to the health of both mother and baby. Therefore, it is necessary to collect past electronic medical history data to provide more accurate and comprehensive information for assessing the pregnancy risks of pregnant women.

[0003] In existing technologies, pregnancy risk assessment methods typically rely on basic vital signs data and medical history from previous electronic medical histories to make risk judgments based on experience and knowledge. However, due to individual differences among pregnant women and complex environmental and lifestyle factors, the analysis of complex data in the medical history is easily affected by different non-essential vital signs parameters, resulting in reduced efficiency and poor accuracy in assessing pregnancy risk. Summary of the Invention

[0004] To address the technical problem of reduced efficiency and accuracy in assessing pregnancy risks due to the influence of various non-essential vital signs, this invention aims to provide a pregnancy risk assessment system based on prior electronic medical history. The specific technical solution adopted is as follows:

[0005] This invention proposes a pregnancy risk assessment system for pregnant women based on their previous electronic medical history. This system is configured to execute a pregnancy risk assessment method based on previous electronic medical history, the method comprising:

[0006] Data on multiple vital signs of pregnant women during different examinations were obtained sequentially.

[0007] For different examination items, based on the distribution of detection data for each vital sign parameter at different times, the abnormality coefficient of each vital sign parameter at each time is obtained, and important vital sign parameters are screened out.

[0008] For any important vital sign parameter, obtain the fitting curve of the corresponding detection data at all times. Based on the distribution of the fitting data in the local range of each time on the fitting curve and the abnormality coefficient corresponding to each time, obtain the detection state abnormality degree of the local range at each time. Based on the time difference between adjacent times and the detection state abnormality degree of the local range at the corresponding time, obtain the detection state change index between adjacent times.

[0009] Based on the changing trends of the detection state change indicators between all adjacent time points, the normal time period and the initial abnormal time period are obtained; based on the time distribution characteristics between each initial abnormal time period and the adjacent normal time period, and the corresponding detection state change indicators within the initial abnormal time period, the abnormality degree state coefficient of each initial abnormal time period is obtained, and the actual abnormal time period is selected.

[0010] Based on the real-time detection data and corresponding detection status change indicators, the detection data for the next time moment is obtained; the detection data for each important vital sign parameter in the neighborhood of the real-time moment or the neighborhood of the next time moment are analyzed to obtain the abnormality degree state coefficient for each actual abnormal time period, and the high abnormal time period is screened out.

[0011] Furthermore, the method for obtaining the anomaly coefficient includes:

[0012] Obtain the preset normal data range for each vital sign parameter;

[0013] The ratio of the detection data of each vital sign parameter at each time step to the maximum value within the preset normal data range is obtained as the abnormality coefficient of each vital sign parameter at each time step.

[0014] Furthermore, the method for obtaining the important vital signs parameters includes:

[0015] For any examination item, obtain the number of times each vital sign parameter's data does not fall within the preset normal data range at different times, obtain the number of times all vital sign parameters' data does not fall within the preset normal data range at different times, and obtain the ratio of the number of abnormalities for each vital sign parameter and all vital sign parameters, as the abnormal frequency of each vital sign parameter.

[0016] Based on the fluctuation degree and frequency of the abnormal coefficient of each vital sign parameter at different times, the importance index of each vital sign parameter is obtained. The fluctuation degree and frequency of the abnormal coefficient are positively correlated with the importance index.

[0017] If the importance index of a vital sign parameter is greater than or equal to a preset importance threshold, the corresponding vital sign parameter will be regarded as an important vital sign parameter.

[0018] Furthermore, the method for obtaining the anomaly degree of the detection state includes:

[0019] Based on the fluctuation characteristics of the data differences between adjacent time points within a local range at each time point, and the maximum value of the anomaly coefficients at all corresponding time points, the detection state anomaly degree of the local range at the corresponding time point is obtained. The fluctuation characteristics of the data differences are negatively correlated with the detection state anomaly degree, and the maximum value of the anomaly coefficients is positively correlated with the detection state anomaly degree.

[0020] Furthermore, the method for obtaining the detection state change index includes:

[0021] The time difference between adjacent moments is negatively correlated and normalized to serve as the time influence coefficient.

[0022] The time influence coefficient between adjacent time points is fused with the maximum value of the detection state anomaly degree in the local range of all corresponding time points to obtain the detection state change index between adjacent time points.

[0023] Furthermore, obtaining the normal time period and the initial abnormal time period includes:

[0024] Obtain the detection state change index sequence between all adjacent time points according to the time sequence; perform differential processing on the detection state change index sequence to obtain the differential sequence of the detection state change index sequence;

[0025] If a target element appears in the difference sequence, and the target element is greater than a preset difference threshold, and the difference between the corresponding element and the next adjacent element is less than or equal to a preset difference threshold, the continuous time range of the target element and the next adjacent element corresponding to the detection state change index in the detection state change index sequence is taken as the initial abnormal time period; other continuous time ranges other than the initial abnormal time period are taken as normal time periods.

[0026] Furthermore, the method for obtaining the anomaly degree state coefficient includes:

[0027] The sum of the durations of the adjacent normal time periods before and after each initial abnormal time period is obtained and used as the first sum value;

[0028] Obtain the ratio of the duration of each initial abnormal time period to the first sum value, and normalize it to obtain the first abnormality coefficient;

[0029] The maximum value of the corresponding detection state change index within the initial abnormal time period and the first abnormal coefficient are fused together to form the abnormality degree state coefficient for each initial abnormal time period.

[0030] Furthermore, the method for obtaining the actual abnormal time period includes:

[0031] If the abnormality level coefficient of the initial abnormal time period is greater than the preset coefficient threshold, the corresponding initial abnormal time period will be taken as the actual abnormal time period.

[0032] Furthermore, the method for acquiring the detection data at the next moment includes:

[0033] For any important vital sign parameter, the real-time detection data and the corresponding detection state change index are fused to obtain the detection data at the next moment.

[0034] Furthermore, the method for obtaining the high-anomaly time period includes:

[0035] The maximum value of the anomaly degree state coefficient within the neighborhood of the real-time time period is selected as the reference anomaly degree.

[0036] Within the neighborhood of the next time moment, the ratio of the anomaly degree state coefficient of each actual anomaly time period to the reference anomaly degree is obtained, which is used as the anomaly weight of each actual anomaly time period.

[0037] If the abnormal weight of the actual abnormal time period is greater than the preset weight threshold, the corresponding actual abnormal time period will be regarded as a high abnormal time period.

[0038] The present invention has the following beneficial effects:

[0039] This invention, for different examination items, obtains the abnormality coefficient of each vital sign parameter at each time point based on the distribution of detection data at different times, and screens out important vital sign parameters, particularly those that play a crucial role in diagnosis. For any important vital sign parameter, it obtains a fitting curve of the corresponding detection data at all times, reflecting the trend of the vital sign parameter's change over time, which helps identify potential patterns of change. Based on the distribution of the fitted data within a local range at each time point on the fitting curve, and the corresponding abnormality coefficient at each time point, it obtains the abnormality degree of the detection state within a local range at each time point, allowing for a more detailed assessment of the detection state at each time point. Based on the time difference between adjacent time points and the abnormality degree of the detection state within a local range at the corresponding time point, it obtains an index of the change in the detection state between adjacent time points, which can quantitatively assess the rate and trend of change of vital sign parameters, helping to identify rapid changes or... Abnormal fluctuations; by analyzing the changing trends of detection status indicators between all adjacent time points, normal time periods and initial abnormal time periods are obtained, which helps to narrow the scope of abnormal detection and improve the accuracy and efficiency of diagnosis; based on the temporal distribution characteristics between each initial abnormal time period and adjacent normal time periods, and the corresponding detection status change indicators within the initial abnormal time period, the abnormality degree state coefficient of each initial abnormal time period is obtained, and the actual abnormal time periods are screened out, allowing for a more accurate assessment of the severity of each abnormal time period; based on the detection data at real time and the corresponding detection status change indicators, the detection data for the next time period is obtained, allowing for advance understanding of possible changes in vital signs parameters; by analyzing the detection data of each important vital sign parameter within the neighborhood range of the real time period or the neighborhood range of the next time period, the abnormality degree state coefficient corresponding to each actual abnormal time period is obtained, and high-abnormal time periods are screened out. This invention improves the accuracy of pregnancy risk assessment by obtaining important vital sign parameters and accurate distribution of abnormal states. Attached Figure Description

[0040] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 A flowchart illustrating a method for assessing pregnancy risk in pregnant women based on prior electronic medical history, provided as an embodiment of the present invention;

[0042] Figure 2 A flowchart illustrating a method for obtaining important vital signs parameters according to an embodiment of the present invention;

[0043] Figure 3 This is a flowchart illustrating a method for obtaining anomaly degree state coefficients according to an embodiment of the present invention. Detailed Implementation

[0044] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a method for assessing pregnancy risk in pregnant women based on prior electronic medical history, proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0045] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0046] The following description, in conjunction with the accompanying drawings, details a specific scheme for a method of assessing pregnancy risk in pregnant women based on prior electronic medical history, provided by this invention.

[0047] Please see Figure 1 The diagram illustrates a flowchart of a method for assessing pregnancy risk in pregnant women based on prior electronic medical history, according to an embodiment of the present invention. The specific method includes:

[0048] Step S1: Obtain the detection data of multiple vital signs parameters of pregnant women in different examination items according to the time sequence.

[0049] In embodiments of the present invention, to improve data analysis efficiency and the accuracy of pregnancy risk diagnosis, the analysis of vital signs parameters is required during the analysis of the pregnant woman's electronic medical history to avoid the influence of a large amount of useless data. Firstly, the electronic medical history form includes tests such as complete blood count and clinical laboratory tests. Each test includes multiple vital signs parameters, such as white blood cell count, red blood cell count, and platelet count. The electronic medical history forms need to be sorted according to the chronological order of the pregnant woman's condition. Sorting the test items in different forms in a consistent order helps reduce information omissions and facilitates the acquisition and analysis of the test data for each vital sign parameter. The test data for multiple vital signs parameters in different test items are acquired chronologically.

[0050] It should be noted that, in the embodiments of the present invention, during each test of a pregnant woman, the implementer can obtain the corresponding next moment in advance according to the specific circumstances.

[0051] Step S2: For different examination items, based on the distribution of detection data for each vital sign parameter at different times, obtain the abnormality coefficient of each vital sign parameter at each time, and screen out important vital sign parameters.

[0052] The test results of pregnant women vary at different times. By analyzing the distribution of test data for each vital sign parameter at different times, we can reflect the data fluctuations at different times and quantify the degree of abnormality of the vital sign parameter. The greater the degree of abnormality, the more important the vital sign parameter needs to be monitored, which helps in the subsequent analysis of abnormal conditions. For different test items, based on the distribution of test data for each vital sign parameter at different times, we can obtain the abnormality coefficient of each vital sign parameter at each time and screen out the important vital sign parameters.

[0053] Preferably, in one embodiment of the present invention, the method for obtaining the anomaly coefficient includes:

[0054] Obtain the preset normal data range for each vital sign parameter;

[0055] The ratio of the detection data of each vital sign parameter at each time step to the maximum value within the preset normal data range is obtained as the abnormality coefficient of each vital sign parameter at each time step.

[0056] It should be noted that, in one embodiment of the present invention, the preset normal data range can be used to quantify the degree of deviation between the detection data of vital signs at different times and the normal data distribution. The preset normal data range of vital signs is obtained in advance by the implementer based on relevant professional data. For example, the normal data range of white blood cell count in adults is 4.0 × 10⁻⁶. 9 ~10.0×10 9 per L.

[0057] Preferably, in one embodiment of the present invention, the method for obtaining important vital signs parameters is described in [reference needed]. Figure 2 It shows a flowchart of a method for obtaining important vital signs parameters, including:

[0058] Step S201: For any examination item, obtain the number of times each vital sign parameter's test data does not fall within the preset normal data range at different times, obtain the number of times all vital sign parameters' test data does not fall within the preset normal data range at different times, and obtain the ratio of the number of abnormalities for each vital sign parameter and all vital sign parameters, as the abnormal frequency of each vital sign parameter.

[0059] By analyzing the frequency of abnormalities, we can identify which vital signs are more likely to deviate from the normal range. The higher the frequency of abnormalities, the greater the likelihood that the corresponding vital signs are abnormal.

[0060] Step S202: Based on the fluctuation degree of the abnormal coefficient of each vital sign parameter at different times and the abnormal frequency, obtain the importance index of each vital sign parameter. The fluctuation degree of the abnormal coefficient and the abnormal frequency are positively correlated with the importance index.

[0061] The degree of fluctuation in the anomaly coefficient can indicate the dispersion of abnormal data in vital signs at different times. The greater the fluctuation, the more unstable the data changes, the more attention is needed, and the greater the importance of the indicator. The higher the frequency of anomalies, the more times the vital signs data are abnormal, which has a greater impact on pregnant women and postpartum women, and the greater the importance of the indicator.

[0062] It should be noted that, in one embodiment of the present invention, the degree of fluctuation can be represented by calculating the variance. The larger the variance, the greater the degree of fluctuation, and the smaller the variance, the smaller the degree of fluctuation. In other embodiments of the present invention, the degree of fluctuation can also be represented by standard deviation, range, etc. The specific means are well known to those skilled in the art and will not be described in detail here.

[0063] In one embodiment of the present invention, the formula for the importance index for any inspection item is expressed as:

[0064] ;

[0065] in, Indicates the first Importance indicators of individual characteristics; Indicates the first The degree of fluctuation of the abnormal coefficients of individual trait parameters at all times; Indicates the first The number of times that an individual trait parameter data does not fall within the preset normal data range at all times; This indicates the number of abnormal data points for all vital signs at all times, where the measured data do not fall within the preset normal range. Represents the normalization function; Indicates the first The ratio of individual trait parameters to the number of abnormalities for all trait parameters is used as the abnormality frequency.

[0066] Step S203: If the importance index of the vital signs parameter is greater than or equal to the preset importance threshold, the corresponding vital signs parameter is regarded as an important vital signs parameter.

[0067] It should be noted that, in one embodiment of the present invention, the preset importance threshold is 0.4. In other embodiments of the present invention, the size of the preset importance threshold can be set according to specific circumstances, and will not be limited or elaborated here.

[0068] Based on this, different examination items were analyzed to obtain all important vital signs parameters.

[0069] Step S3: For any important vital sign parameter, obtain the fitting curve of the corresponding detection data at all times. Based on the distribution of the fitting data in the local range of each time on the fitting curve and the abnormality coefficient at each time, obtain the detection state abnormality degree of the local range at each time. Based on the time difference between adjacent times and the detection state abnormality degree of the local range at the corresponding time, obtain the detection state change index between adjacent times.

[0070] To gain a more intuitive understanding of the changing trends and patterns of the detection data at different times, for any important vital sign parameter, we obtain the fitting curves of the corresponding detection data at all times.

[0071] It should be noted that, in one embodiment of the present invention, the corresponding fitting curve is obtained by performing least squares fitting on the detection data at all times, i.e., the fitting data with time on the horizontal axis and vital sign parameters on the vertical axis; in other embodiments of the present invention, curve fitting can also be performed by existing fitting algorithms such as polynomial fitting and spline interpolation; the specific means are well known to those skilled in the art and will not be described in detail here.

[0072] The distribution of fitted data reflects the volatility and stability of the fitted data. A relatively concentrated and stable distribution of data in a local area indicates that the data state is relatively normal. Conversely, if the data distribution is discrete or exhibits abnormal fluctuations, it may indicate that the data is in an abnormal state. Since the data state is relatively stable even when the abnormal state persists, by introducing an anomaly coefficient, we can more accurately identify which moments in the data are in an abnormal state. Based on the distribution of fitted data in a local area at each moment on the fitted curve, and the anomaly coefficient corresponding to each moment, we can obtain the degree of anomaly detected in the local area at each moment.

[0073] Preferably, in one embodiment of the present invention, the method for obtaining the state anomaly degree includes:

[0074] Based on the fluctuation characteristics of the data differences between adjacent time points within the local range of each time point, and the maximum value of the abnormal coefficient in all corresponding time points, the detection state abnormality degree of the local range at the corresponding time point is obtained. The fluctuation characteristics of the data differences are negatively correlated with the detection state abnormality degree, and the maximum value of the abnormal coefficient is positively correlated with the detection state abnormality degree.

[0075] Among them, the greater the fluctuation characteristics of data differences, the more dissimilar the data differences, the more unstable the state performance, the larger the anomaly coefficient, and the greater the degree of anomaly detection.

[0076] It should be noted that, in one embodiment of the present invention, the mean of the ratios between the differences of all adjacent data within a local range at a given time is used as the level of difference change; wherein, to avoid the denominator of the formula being 0 and the formula being meaningless, a manually set threshold, such as 0.01, is added to the denominator when calculating the ratio; the difference between the level of difference change and the positive integer 1 is calculated, and then... The function is normalized to obtain the state stability coefficient, which reflects the fluctuation characteristics. That is, the larger the state stability coefficient, the closer the difference between the level of difference change and the positive integer 1 is, the smaller the fluctuation characteristics of the data difference, indicating that the results of multiple tests are relatively consistent.

[0077] In one embodiment of the present invention, the formula for detecting the degree of state anomaly within a local range at any given time is expressed as:

[0078] ;

[0079] in, This indicates the degree of anomaly in the detection state within a local area at each time step; Indicates the number of moments within a local time range; Indicates the first Time and the The data differences in the fitted data between time points; Indicates the first Time and the The data differences in the fitted data between time points; This represents the maximum value of the anomaly coefficient within a local range at a given time. It represents the first derivative of the logistic function.

[0080] In the formula for detecting the degree of state abnormality, Indicates the first Time and the The data difference between fitted data at time points and the first time point Time and the The ratio of the differences in the fitted data between time points; This represents the mean of all ratios within a local range at a given time point, indicating the level of variation. A higher level of variation results in a larger mean and a larger ratio of data differences, signifying greater instability in the data across time points. When normalizing, the less similar the data differences are, the greater the fluctuation of data changes and the lower the detection anomaly degree. Conversely, the closer the data differences are, the smaller the fluctuation of data changes, the larger the maximum value of the anomaly coefficient and the higher the detection anomaly degree.

[0081] It should be noted that, in one embodiment of the present invention, the local time range is the range formed by each time point as the main point and the two adjacent time points, that is, traversing the fitted curve and analyzing the different local time ranges obtained in turn; in other embodiments of the present invention, the size of the local time range can be set according to the specific situation, which will not be limited or described here.

[0082] The time span of examinations for pregnant women varies. If there are no abnormalities or the degree of abnormality is low, the interval between the next examination may be long, possibly due to other conditions not being detected. By analyzing the magnitude of the time difference between adjacent moments, the finer details of the changes in the state can be captured. The smaller the time difference, the more attention can be paid to subtle changes in the degree of abnormality in the test state, and the greater the impact on the degree of abnormality in the test state. The larger the time difference, the more likely some important changes will be missed, and the smaller the impact on the degree of abnormality in the test state. Based on the time difference between adjacent moments and the degree of abnormality in the test state in the corresponding local area, an index of the change in the test state between adjacent moments can be obtained.

[0083] Preferably, in one embodiment of the present invention, the method for obtaining the detection state change index includes:

[0084] The time difference between adjacent moments is negatively correlated and normalized to serve as the time influence coefficient.

[0085] The time influence coefficient between adjacent time points is fused with the maximum value of the detection state anomaly degree in the local range of all corresponding time points to obtain the detection state change index between adjacent time points.

[0086] In one embodiment of the present invention, for the first Time and the At any given time, the formula for the index of state change is expressed as:

[0087] ;

[0088] in, Indicators representing changes in detection status between adjacent time points; Indicates the first Time and the The time difference between moments; This represents the maximum time difference between all adjacent moments; This represents the maximum value of the detected state anomaly degree for a local range across all time points; This represents the maximum and minimum normalization functions.

[0089] In the formula for detecting state change indicators, This represents the ratio of the maximum time difference to the time difference between adjacent moments, and is normalized. That is, the time difference between adjacent moments is negatively normalized to obtain the time influence coefficient. The larger the time influence coefficient, the larger the time difference between adjacent moments, the smaller the impact of the detection data on the pregnant woman's condition, the smaller the abnormality of the detection condition, and the smaller the change index of the detection condition.

[0090] Step S4: Based on the changing trends of the detection state change indicators between all adjacent time points, obtain the normal time period and the initial abnormal time period; based on the time distribution characteristics between each initial abnormal time period and the adjacent normal time period, and the corresponding detection state change indicators within the initial abnormal time period, obtain the abnormality degree state coefficient of each initial abnormal time period, and filter out the actual abnormal time periods.

[0091] The detection status change index reflects the changing trend of the pregnant and postpartum woman's status at different times. By dividing the normal time period, it is helpful to analyze the status changes in each time period in more detail. Based on the changing trend of the detection status change index between different adjacent times, the normal time period and the initial abnormal time period can be obtained.

[0092] Preferably, in one embodiment of the present invention, the method for obtaining the normal time period and the initial abnormal time period includes:

[0093] Obtain the detection state change index sequence between all adjacent time points according to the time sequence; perform differential processing on the detection state change index sequence to obtain the differential sequence of the detection state change index sequence;

[0094] If a target element appears in the difference sequence, and the target element is greater than the preset difference threshold, and the difference between the corresponding element and the next adjacent element is less than or equal to the preset difference threshold, the continuous time range of the detection state change index corresponding to each target element and the next adjacent element on the detection state change index sequence is taken as the initial abnormal time period; other continuous time ranges other than the initial abnormal time period are taken as normal time periods.

[0095] It should be noted that, in one embodiment of the present invention, the preset difference threshold is 0.7 and the preset difference threshold is 0.1; in other embodiments of the present invention, the preset difference threshold and the preset difference threshold can be set by the implementer according to the specific circumstances, and are not limited or elaborated here.

[0096] For example, if there exists a detection state change index sequence of 0.1, 0.2, 0.2, 0.9, 0.3, 0.2, with adjacent times of 12, 23, 35, 57, 78, 89, the resulting difference sequence is 0.1, 0, 0.7, 0.6, 0.1. If the third element meets the condition of being greater than a preset difference threshold and the difference between it and the next adjacent element is less than or equal to a preset difference threshold, then the third element is the target element. The third and fourth elements detect state change indices 0.2, 0.9, and 0.3 in the detection state change index sequence. The time range of the adjacent times 35, 57, and 78 is 3-8, which is considered the initial abnormal time period. Other consecutive time ranges of 1-3 or 8-9 are considered normal time periods.

[0097] Pregnant women's bodies are generally in a stable state, and the similarity between their state before and after an abnormal state is relatively high. Analyzing the temporal distribution characteristics between each initial abnormal time period and the adjacent normal time period can reflect the persistence and trend of the abnormality. Detecting state change indicators is an important indicator for assessing state changes within the abnormal time period. The larger the detected state change indicator, the greater the possibility of a state change and the greater the degree of influence from the abnormality. Comprehensive analysis provides a more complete understanding of the actual situation of the abnormal time period. Based on the temporal distribution characteristics between each initial abnormal time period and the adjacent normal time period, as well as the corresponding detected state change indicators within the initial abnormal time period, the abnormality degree state coefficient of each initial abnormal time period is obtained, and the actual abnormal time periods are screened out.

[0098] Preferably, in one embodiment of the present invention, the method for obtaining the anomaly degree state coefficient is described in [reference needed]. Figure 3 It shows a flowchart of a method for obtaining anomaly degree state coefficients, including:

[0099] Step S301: Obtain the sum of the durations of the adjacent normal time periods before and after each initial abnormal time period, as the first sum value; obtain the ratio of the duration of each initial abnormal time period to the first sum value, and normalize it, as the first abnormal coefficient.

[0100] By measuring duration, the degree of change can be captured. The ratio of the duration of each abnormal time period to the first sum reflects the comprehensive performance of the abnormal time period in terms of its temporal relationship with adjacent normal time periods. The shorter the duration of the abnormal time period, the more likely it is to be a short-term anomaly, and the smaller the first anomaly coefficient.

[0101] Step S302: Combine the maximum value of the corresponding detection state change index within the initial abnormal time period with the first abnormal coefficient to form the abnormality degree state coefficient for each initial abnormal time period.

[0102] In some embodiments of the present invention, fusion is performed by addition or multiplication. The specific means are well known to those skilled in the art and are not limited or described in detail here.

[0103] In one embodiment of the present invention, the formula for the anomaly degree state coefficient is expressed as:

[0104] ;

[0105] in, Indicates the first The anomaly degree state coefficient for each initial abnormal time period; Indicates the first The duration of the initial abnormal period; Indicates the first The duration of the preceding normal period adjacent to each initial abnormal period; Indicates the first The duration of the next normal period adjacent to the initial abnormal period; Indicates the first The maximum value of the corresponding detection status change index within each initial abnormal time period; This represents a non-linear activation function.

[0106] In the formula for the state coefficient of the degree of abnormality, Indicates the first sum; Indicates the calculation of the first The ratio of the duration of the initial abnormal time period to the first sum value; Indicates the first The ratio of the duration of the initial abnormal time period to the first sum is normalized, which is the first abnormality coefficient. The larger the ratio, the stronger the abnormality. The smaller the sum of the durations of the preceding and following normal periods of an initial abnormal period, the better. The longer the initial abnormal time period, the larger the first abnormality coefficient and the larger the abnormality degree state coefficient; the... The larger the maximum value of the corresponding detection state change index within an initial abnormal time period, the higher the degree of abnormal influence and the smaller the abnormality state coefficient.

[0107] Considering that improper diet and excessive salt intake may cause short-term parameter abnormalities, but such abnormalities are of no value to pregnancy risk, screening for the actual time period of abnormality can help improve the efficiency of detection.

[0108] Preferably, in one embodiment of the present invention, the method for obtaining the actual abnormal time period includes:

[0109] If the abnormality level coefficient of an abnormal time period is greater than the preset coefficient threshold, the corresponding initial abnormal time period will be taken as the actual abnormal time period.

[0110] It should be noted that, in one embodiment of the present invention, the preset coefficient threshold is 0.4; in other embodiments of the present invention, the size of the preset coefficient threshold can be set according to specific circumstances, and will not be limited or elaborated here.

[0111] Step S5: Based on the real-time detection data and the corresponding detection status change indicators, obtain the detection data for the next time step; analyze the detection data of each important vital sign parameter in the neighborhood range of the real-time time step or the neighborhood range of the next time step to obtain the abnormality degree state coefficient corresponding to each actual abnormal time period, and filter out the high abnormal time periods.

[0112] By monitoring real-time changes in a patient's vital signs, their health status can be understood promptly. Monitoring status change indicators provides information on the trends of these vital signs, allowing for prediction of future test data and thus providing advance knowledge of the patient's potential health condition. Based on real-time test data and corresponding status change indicators, the test data for the next moment can be obtained.

[0113] Preferably, in one embodiment of the present invention, the method for acquiring the detection data at the next moment includes:

[0114] For any important vital sign parameter, the real-time detection data and the corresponding detection state change index are fused to obtain the detection data at the next moment.

[0115] It should be noted that in some embodiments of the present invention, fusion can be achieved by addition or multiplication. The specific means are well known to those skilled in the art and will not be described in detail here.

[0116] The abnormality degree state coefficient can quantify the degree of abnormality of vital signs and parameters within an abnormal time period, thereby providing a more accurate understanding of the patient's health status. By analyzing the detection data of each important vital sign and parameter within its neighborhood at a real time or within its neighborhood at the next time period, the abnormality degree state coefficient corresponding to each actual abnormal time period can be obtained. Highly abnormal time periods can be screened out, which helps to understand the patient's health status in a timely manner and improves the accuracy and efficiency of diagnosis.

[0117] Preferably, in one embodiment of the present invention, the method for obtaining high anomaly time periods includes:

[0118] The maximum value of the anomaly degree state coefficient within the neighborhood of the real-time time period is selected as the reference anomaly degree.

[0119] Within the neighborhood of the next time step, the ratio of the anomaly degree state coefficient to the reference anomaly degree for each actual anomaly time period is obtained, which is used as the anomaly weight for each actual anomaly time period.

[0120] If the abnormal weight of the actual abnormal time period is greater than the preset weight threshold, the corresponding actual abnormal time period will be regarded as a high abnormal time period.

[0121] It should be noted that, in one embodiment of the present invention, the preset weight threshold is 0.5; in other embodiments of the present invention, the size of the preset weight threshold can be set according to specific circumstances, and will not be limited or elaborated here.

[0122] It should be noted that, in one embodiment of the present invention, the method for obtaining the neighborhood range is to take each moment as a reference and form a range with all historical moments; in other embodiments of the present invention, the size of the neighborhood range can be specifically set according to the specific situation, and will not be limited or described in detail here.

[0123] Based on this, by analyzing all important vital signs and parameters and identifying periods of high abnormality, a detailed analysis of the pregnant woman's condition can be conducted, improving the accuracy of analyzing abnormal conditions in pregnant women and increasing detection efficiency.

[0124] In summary, this invention obtains the degree of abnormality in the detection state of a local area at each moment for any important vital sign parameter; combines the time difference between adjacent moments to obtain the index of changes in the detection state between adjacent moments; performs temporal feature analysis on the normal time period and the initial abnormal time period, and combines the corresponding detection state change index within the initial abnormal time period to obtain the abnormality degree state coefficient for each initial abnormal time period, and filters out the actual abnormal time periods; analyzes the detection data of each important vital sign parameter in the neighborhood range of the real time moment or the neighborhood range of the next moment moment to obtain the abnormality degree state coefficient for each actual abnormal time period, and filters out the high-abnormal time periods. This invention improves the accuracy of pregnancy risk assessment by obtaining important vital sign parameters and accurate distribution of abnormal states.

[0125] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0126] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A pregnancy risk assessment system for pregnant women based on their past electronic medical history, characterized in that, The system is configured to perform the following steps: Data on multiple vital signs of pregnant women during different examinations were obtained sequentially. For different examination items, based on the distribution of the detection data of each vital sign parameter at different times, the abnormality coefficient of each vital sign parameter at each time is obtained, and important vital sign parameters are screened out; the abnormality coefficient of each vital sign parameter at each time is the ratio of the detection data of each vital sign parameter at each time to the maximum value within the preset normal data range. The important vital signs parameters are those whose importance index is greater than or equal to a preset importance threshold; the formula for the importance index is as follows: ； in, Indicates the first Importance indicators of individual characteristics; Indicates the first The degree of fluctuation of the abnormal coefficients of individual trait parameters at all times; Indicates the first The number of times that an individual trait parameter data does not fall within the preset normal data range at all times; This indicates the number of abnormal data points for all vital signs at all times, where the measured data do not fall within the preset normal range. Represents the normalization function; Indicates the first The ratio of individual trait parameters to the number of abnormalities for all trait parameters is used as the abnormality frequency. For any important vital sign parameter, obtain the fitting curve of the corresponding detection data at all time points. Based on the distribution of the fitting data within a local range at each time point on the fitting curve, and the anomaly coefficient corresponding to each time point, obtain the detection state anomaly degree of the local range at each time point. Based on the time difference between adjacent time points and the detection state anomaly degree of the corresponding local range at each time point, obtain the detection state change index between adjacent time points. The formula for the detection state anomaly degree is expressed as: ； in, This indicates the degree of anomaly in the detection state within a local area at each time step; Indicates the number of moments within a local time range; Indicates the first Time and the The data differences in the fitted data between time points; Indicates the first Time and the The data differences in the fitted data between time points; This represents the maximum value of the anomaly coefficient within a local range at a given time. The first derivative of the logistic function; The formula for the index of detected state change is expressed as: ； in, Indicators representing changes in detection status between adjacent time points; Indicates the first Time and the The time difference between moments; This represents the maximum time difference between all adjacent moments; This represents the maximum value of the detected state anomaly degree for a local range across all time points; Represents the maximum and minimum normalization functions; Based on the changing trends of the detected state change indicators between all adjacent time points, the normal time period and the initial abnormal time period are obtained. Based on the temporal distribution characteristics between each initial abnormal time period and the adjacent normal time period, and the corresponding detected state change indicators within the initial abnormal time period, the abnormality degree state coefficient for each initial abnormal time period is obtained, and the actual abnormal time periods are selected. The formula for the abnormality degree state coefficient is expressed as: ； in, Indicates the first The anomaly degree state coefficient for each initial abnormal time period; Indicates the first The duration of the initial anomaly; Indicates the first The duration of the preceding normal period adjacent to each initial abnormal period; Indicates the first The duration of the next normal period adjacent to the initial abnormal period; Indicates the first The maximum value of the corresponding detection status change index within each initial abnormal time period; Represents a nonlinear activation function; Based on the real-time detection data and corresponding detection status change indicators, the detection data for the next time moment is obtained; the detection data for each important vital sign parameter in the neighborhood of the real-time moment or the neighborhood of the next time moment are analyzed to obtain the abnormality degree state coefficient for each actual abnormal time period, and the high abnormal time period is screened out.

2. The pregnancy risk assessment system for pregnant women based on past electronic medical history as described in claim 1, characterized in that, The acquisition of the normal time period and the initial abnormal time period includes: Obtain the detection state change index sequence between all adjacent time points according to the time sequence; perform differential processing on the detection state change index sequence to obtain the differential sequence of the detection state change index sequence; If a target element appears in the difference sequence, and the target element is greater than a preset difference threshold, and the difference between the corresponding element and the next adjacent element is less than or equal to a preset difference threshold, the continuous time range of the target element and the next adjacent element corresponding to the detection state change index in the detection state change index sequence is taken as the initial abnormal time period; other continuous time ranges other than the initial abnormal time period are taken as normal time periods.

3. The pregnancy risk assessment system for pregnant women based on past electronic medical history as described in claim 1, characterized in that, The method for obtaining the actual abnormal time period includes: If the abnormality level coefficient of the initial abnormal time period is greater than the preset coefficient threshold, the corresponding initial abnormal time period will be taken as the actual abnormal time period.

4. The pregnancy risk assessment system for pregnant women based on past electronic medical history as described in claim 1, characterized in that, The method for obtaining the detection data at the next moment includes: For any important vital sign parameter, the real-time detection data and the corresponding detection state change index are fused to obtain the detection data at the next moment.

5. A pregnancy risk assessment system for pregnant women based on past electronic medical history as described in claim 1, characterized in that, The method for obtaining the high-anomaly time period includes: The maximum value of the anomaly degree state coefficient within the neighborhood of the real-time time period is selected as the reference anomaly degree. Within the neighborhood of the next time moment, the ratio of the anomaly degree state coefficient of each actual anomaly time period to the reference anomaly degree is obtained, which is used as the anomaly weight of each actual anomaly time period. If the abnormal weight of the actual abnormal time period is greater than the preset weight threshold, the corresponding actual abnormal time period will be regarded as a high abnormal time period.

Images