An Obstetric Real-time Nursing Optimization Method and System Based on Big Data

Abstract

The present invention relates to a real-time obstetric nursing optimization method and system based on big data, belonging to the field of medical nursing. Taking the static indicators of obstetric nursing objects as constants and the dynamic indicators as variables, multiple abnormal nursing logs are retrieved; based on the set of abnormal dynamic indicators, the central dynamic indicators with the number of occurrences of dynamic indicators in the abnormal nursing logs greater than or equal to the number threshold and the marginal dynamic indicators with the number of occurrences less than the number threshold are statistically analyzed; the deviation vector of the monitoring values of the central dynamic indicators and the proportion of the number of deviation indicators of the monitoring values of the marginal dynamic indicators are statistically analyzed; the probability of abnormal nursing that simultaneously meets the static indicators, the deviation vector and the proportion of the number of deviation indicators is predicted; when the probability is greater than or equal to the probability threshold, the nursing end is reminded to perform nursing optimization, solving the technical problem that the individual differences among pregnant women are ignored during real-time obstetric nursing, resulting in insufficient accuracy and efficiency of real-time nursing.

Description

Technical Field

[0001] The present invention relates to the field of medical care, and particularly to an optimization method and system for real-time obstetric care based on big data. Background Art

[0002] In the traditional real-time obstetric care model, big data statistics are widely used to predict and judge the abnormal care risks of pregnant women in different states. However, this traditional data statistics method has a significant problem of insufficient individualization. It usually directly collects the real-time status of all indicators of pregnant women and evaluates based on the probability of these statuses being abnormal in historical data. But this method ignores the differences between individual pregnant women. Different pregnant women have different tolerance levels for different physiological indicators. Specifically, the physiological state of pregnant women is affected by various factors, including age, weight, pregnancy stage, health status, etc. These factors result in different sensitivities and response degrees of pregnant women to different indicators. The traditional data statistics method does not fully consider these individual differences, but uses a unified standard and threshold for all pregnant women for evaluation, which is obviously inaccurate and lacks individualization. In addition, the traditional data statistics method also has certain limitations in the judgment of abnormal situations. It usually only focuses on whether the indicators exceed the normal range and lacks in-depth analysis of the change trend of indicators and abnormal situations. This simple binary classification judgment method is prone to false alarms and missed alarms, affecting the accuracy and efficiency of real-time obstetric care. Summary of the Invention

[0003] In view of the technical problem in the prior art that the individual differences between pregnant women are ignored during real-time obstetric care, resulting in insufficient accuracy and efficiency of real-time care, the present invention provides an optimization method and system for real-time obstetric care based on big data to solve this problem.

[0004] The technical solution of the present invention to solve the above technical problems is as follows:

[0005] In a first aspect, the present invention provides an optimization method for real-time obstetric care based on big data, including: taking the static indicators of the obstetric care object as constants and the dynamic indicators as variables, retrieving multiple abnormal care logs, where any one of the abnormal care logs includes a set of abnormal dynamic indicators; based on the set of abnormal dynamic indicators, counting the central dynamic indicators whose occurrence log count in the multiple abnormal care logs is greater than or equal to the log count threshold, and the marginal dynamic indicators whose occurrence log count is less than the log count threshold; counting the deviation vector of the monitoring values of the central dynamic indicators of the obstetric care object, and the proportion of the deviation index quantity of the monitoring values of the marginal dynamic indicators; predicting the abnormal care probability that simultaneously satisfies the static indicators, the deviation vector, and the proportion of the deviation index quantity; when the abnormal care probability is greater than or equal to the abnormal probability threshold, reminding the nursing end to perform care optimization.

[0006] In a second aspect, the present invention provides a real-time obstetric care optimization system based on big data, comprising: an information retrieval module, configured to retrieve multiple abnormal care logs with static indicators of an obstetric care object as constants and dynamic indicators as variables, wherein any one of the abnormal care logs includes a set of abnormal dynamic indicators; a statistical screening module, configured to, based on the set of abnormal dynamic indicators, statistically screen central dynamic indicators whose occurrence log count in the multiple abnormal care logs is greater than or equal to a log count threshold, and marginal dynamic indicators whose occurrence log count is less than the log count threshold; a deviation statistics module, configured to statistically calculate the deviation vector of the monitored values of the central dynamic indicators of the obstetric care object, and the proportion of the number of deviation indicators of the monitored values of the marginal dynamic indicators; an abnormal prediction module, configured to predict the probability of abnormal care that simultaneously satisfies the static indicators, the deviation vector, and the proportion of the number of deviation indicators; and a care optimization module, configured to, when the probability of abnormal care is greater than or equal to an abnormal probability threshold, remind the care end to perform care optimization.

[0007] The beneficial effects of the present invention are as follows: By distinguishing the static indicators and dynamic indicators of the obstetric care object, and statistically analyzing the central dynamic indicators and marginal dynamic indicators, as well as their deviation situations, based on the abnormal care logs, the probability of abnormal care can be predicted more accurately, so as to timely remind the care end to perform care optimization when needed, effectively improving the pertinence and efficiency of real-time obstetric care. Description of the Drawings

[0008] Figure 1 It is a schematic flow chart of a real-time obstetric care optimization method provided by the present invention.

[0009] Figure 2 It is a schematic structural diagram of a real-time obstetric care optimization system provided by the present invention.

[0010] Description of the reference numerals: Information retrieval module 11, statistical screening module 12, deviation statistics module 13, abnormal prediction module 14, care optimization module 15. Detailed Embodiments

[0011] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts fall within the protection scope of the present invention.

[0012] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of the said features. In the description of the present invention, "a plurality of" means two or more unless otherwise specifically defined.

[0013] In the description of the present invention, the term "for example" is used to mean "serving as an example, illustration, or explanation". Any embodiment described as "for example" in the present invention is not necessarily construed as being more preferred or having more advantages than other embodiments. The following description is provided to enable any person skilled in the art to implement and use the present invention. In the following description, details are set forth for purposes of explanation. It should be understood that those of ordinary skill in the art can recognize that the present invention can be implemented without these specific details. In other instances, well-known structures and processes are not elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is to be accorded the widest scope consistent with the principles and features disclosed herein.

[0014] Embodiment 1:

[0015] As Figure 1 shown, an embodiment of the present invention provides a method for optimizing real-time obstetric care based on big data, which is characterized by including:

[0016] S10: Taking the static indicators of the obstetric care object as constants and the dynamic indicators as variables, retrieving multiple abnormal care logs, where any one of the abnormal care logs includes a set of abnormal dynamic indicators.

[0017] S20: Based on the set of abnormal dynamic indicators, statistically calculate the central dynamic indicators for which the number of occurrence logs of the dynamic indicators in the multiple abnormal care logs is greater than or equal to the log number threshold, and the marginal dynamic indicators for which the number of occurrence logs is less than the log number threshold.

[0018] S30: Statistically calculate the deviation vector of the monitoring values of the central dynamic indicators of the obstetric care object, and the proportion of the number of deviation indicators of the monitoring values of the marginal dynamic indicators.

[0019] S40: Predict the probability of abnormal care that simultaneously satisfies the static indicators, the deviation vector, and the proportion of the number of deviation indicators.

[0020] S50: When the probability of abnormal care is greater than or equal to the abnormal probability threshold, remind the nursing end to optimize the care.

[0021] Exemplarily, in obstetric care work, the health status of the care recipient is usually monitored and evaluated based on specific indicators, which can be classified into two major categories: static indicators and dynamic indicators. Among them, static indicators refer to those that are relatively stable in the short term and are not prone to significant changes. These indicators provide a basic health portrait of the care recipient and help to initially understand their overall condition. Static indicators usually include age range, physical condition, pregnancy status, emotional state number, and psychological endurance rating, etc. And the pregnancy status includes gestational age, number of fetuses, fetal position, and placental position, etc. Dynamic indicators refer to those that change over time, environment, or treatment intervention. These indicators are important bases for monitoring changes in the health status of parturients, such as vital signs (including body temperature, pulse, respiration, blood pressure), postpartum blood loss, lochia condition, pain level, environmental status, etc. In obstetric care practice, the static and dynamic indicators of parturients are usually recorded regularly and a nursing log is generated. When abnormal dynamic indicators occur, these logs will become important bases for identifying and handling potential problems. Through a database or an electronic medical record system, parturients in a specific group can be screened according to static indicators (such as age, pregnancy status, etc.), and then whether there are abnormal values in their dynamic indicators can be further retrieved. For any abnormal nursing log, all abnormal dynamic indicators are sorted out to form a set, that is, the abnormal dynamic indicator set. This set can quickly locate the problem and formulate corresponding nursing plans. For example, if the static indicators of a parturient show that she is an elderly parturient with multiple pregnancies, and abnormal conditions such as elevated body temperature, accelerated pulse, and decreased blood pressure appear in her dynamic indicators, then these abnormal dynamic indicators will indicate that the parturient may have potential risks such as infection and cardiac insufficiency and further evaluation and intervention are needed immediately. To sum up, through the classification and monitoring of static and dynamic indicators, the health status of obstetric care recipients can be understood more comprehensively and accurately, potential problems can be discovered and handled in a timely manner, so as to ensure the safety of mother and baby.

[0022] Furthermore, conduct in-depth statistics and analysis on abnormal dynamic indicators to determine which indicators are the central dynamic indicators with frequent abnormalities and which are the marginal dynamic indicators with occasional abnormalities and relatively minor impacts. Before starting the statistics, a threshold for the number of logs needs to be set. This threshold is determined based on the research purpose, the size of the sample, and the prevalence of abnormal situations. It is used to distinguish between central dynamic indicators (with frequent abnormalities) and marginal dynamic indicators (with occasional abnormalities). For each abnormal dynamic indicator in an abnormal nursing log, it will be statistically analyzed across the entire dataset, and the number of times each indicator appears in all abnormal nursing logs will be recorded. This step is to understand the occurrence frequency of each abnormal dynamic indicator. Subsequently, based on the statistical results, the dynamic indicators are classified into two categories, namely central dynamic indicators and marginal dynamic indicators. The central dynamic indicators appear in multiple abnormal nursing logs with a frequency greater than or equal to the previously set threshold for the number of logs. They indicate that the abnormalities of these indicators are relatively common in a specific group and may be key factors leading to adverse pregnancy outcomes or requiring special attention. The marginal dynamic indicators appear in abnormal nursing logs with a frequency less than the threshold for the number of logs. They may only show abnormalities in individual cases, or the association between their abnormalities and adverse pregnancy outcomes is weak. In summary, through the statistics and analysis of the set of abnormal dynamic indicators, a deeper understanding of the health status of obstetric care recipients can be achieved, providing strong support for formulating personalized care plans and clinical decisions.

[0023] To more meticulously evaluate the health status of the care recipients, not only should we focus on the frequent abnormalities in dynamic indicators (central dynamic indicators), but also on the degree to which the monitored values of these indicators deviate from the normal range, as well as the marginal dynamic indicators that occasionally show abnormalities but may be equally important. First, a normal range needs to be set for each central dynamic indicator (usually based on clinical guidelines, historical data, or expert consensus). The deviation criterion can be the degree to which the monitored value exceeds the normal range, such as standard deviation, percentage, etc. For each abnormal care log containing central dynamic indicators, calculate the degree of deviation of the monitored value of this indicator relative to the normal range. This is usually a vector, containing the direction of deviation (above or below the normal range) and the magnitude (the specific value of the deviation). Subsequently, record the deviation vectors of each central dynamic indicator and conduct summary analysis as needed. For example, statistical quantities such as the average degree of deviation and the maximum degree of deviation can be calculated to understand the overall deviation trend of these indicators under abnormal conditions. For marginal dynamic indicators, check and record which indicators' monitored values deviate from the normal range. The deviation here can be any form of abnormality, such as being higher than, lower than, or not within a specific interval. Summarize the deviations of marginal dynamic indicators in all abnormal care logs and calculate the proportion of the number of deviated indicators to the total number of all marginal dynamic indicators. This proportion reflects the overall deviation degree of marginal dynamic indicators under abnormal conditions. In summary, by analyzing the deviation vectors of central dynamic indicators, we can understand the general trend of these indicators deviating from the normal range under abnormal conditions, which is crucial for formulating targeted care plans and clinical intervention measures. Although the proportion of the number of deviated indicators of marginal dynamic indicators is relatively low, it may reveal some uncommon abnormal conditions or potential risks. These indicators also need to be concerned because they may indicate specific health problems or complications. Combining the deviation vectors of central dynamic indicators and the proportion of the number of deviated indicators of marginal dynamic indicators can obtain a more comprehensive assessment of the health status of obstetric care recipients, which helps to formulate personalized care plans, optimize clinical decisions, and ultimately improve the quality of care and patient satisfaction.

[0024] After that, collect static indicators, deviation vectors, and the proportion of the number of deviation indicators, and use advanced algorithms or models (such as machine learning models) to predict the probability of a patient experiencing an abnormal nursing event. This probability value reflects the likelihood of an adverse nursing outcome for the patient in the current nursing state. Preset an abnormal probability threshold to determine whether the patient's nursing state needs to be optimized. If the predicted abnormal nursing probability is greater than or equal to this threshold, the system will trigger a reminder function and send an alarm to the nursing end. After receiving the reminder, the nursing end can immediately view the patient's detailed information and adjust the nursing plan or take other optimization measures as needed. In summary, the entire process comprehensively evaluates static indicators, deviation vectors, and the proportion of the number of deviation indicators, uses an algorithm to predict the abnormal nursing probability, and reminds the nursing end to optimize when necessary, thereby achieving real-time monitoring and dynamic adjustment of the patient's nursing state, which helps improve the nursing quality and ensure the safety of the patient.

[0025] In a preferred embodiment, based on the set of abnormal dynamic indicators, count the central dynamic indicators with the number of occurrence logs of the dynamic indicators in the multiple abnormal nursing logs being greater than or equal to the log number threshold, and the marginal dynamic indicators with the number of occurrence logs being less than the log number threshold, including:

[0026] Extract the first set of abnormal dynamic indicators of the first abnormal nursing log among the multiple abnormal nursing logs, where the first set of abnormal dynamic indicators is counted in a preset length time zone before the nursing object shows abnormality; count the deviation frequency of the first abnormal dynamic indicators in the first set of abnormal dynamic indicators in the preset length time zone; when the deviation frequency of the first abnormal dynamic indicators is greater than or equal to the deviation frequency threshold, consider the number of occurrence logs of the first abnormal dynamic indicators to increase by one; traverse the dynamic indicators and perform statistics based on the multiple abnormal nursing logs to obtain the central dynamic indicators with the number of occurrence logs being greater than or equal to the log number threshold, and the marginal dynamic indicators with the number of occurrence logs being less than the log number threshold.

[0027] Optionally, select the first one from multiple abnormal care logs for analysis. This log records the changes in dynamic indicators of a certain care recipient within a preset length time zone (e.g., the past 24 hours or 48 hours) before the abnormality occurs. Extract all abnormal dynamic indicators within this time zone to form the first set of abnormal dynamic indicators. Herein, the preset length time zone refers to the time range selected when analyzing abnormal care logs, which is used to observe the changes in dynamic indicators. The first set of abnormal dynamic indicators refers to the set of all abnormal dynamic indicators extracted from the first abnormal care log within the preset length time zone. Subsequently, count the deviation frequency of each indicator in the first set of abnormal dynamic indicators within the preset length time zone. The deviation frequency refers to the number of times a certain dynamic indicator deviates from the normal range (or threshold) within the preset time zone, which is used to measure the abnormality degree of the indicator. For each first abnormal dynamic indicator, compare its deviation frequency with a preset deviation frequency threshold. If the deviation frequency of an indicator is greater than or equal to the threshold, it is considered that this indicator plays an important role in the abnormal care event, so increment the number of log entries in which it appears by one. This means that as long as the deviation frequency of this indicator in any abnormal care log reaches or exceeds the threshold, it is considered to have appeared once in this log. Finally, traverse all dynamic indicators and repeat the above steps (extraction, counting deviation frequency, judgment and counting) until all abnormal care logs are processed. During this process, record the number of times each dynamic indicator appears in all logs. After the statistics are completed, identify the central dynamic indicators and marginal dynamic indicators based on the number of times the dynamic indicators appear. Specifically, indicators with the number of appearances greater than or equal to the log entry threshold are regarded as central dynamic indicators, which frequently appear in multiple abnormal care logs and are closely related to abnormal care events. While indicators with the number of appearances less than the log entry threshold are regarded as marginal dynamic indicators, and their appearance frequency in abnormal care logs is relatively low. Through this process, dynamic indicators closely related to abnormal care events can be effectively extracted from a large number of abnormal care logs, providing strong data support for subsequent prediction of abnormal care probability.

[0028] In a preferred embodiment, predicting the abnormal care probability that simultaneously meets the static indicators, the deviation vector, and the proportion of the number of deviation indicators includes:

[0029] Construct a strict consistency constraint condition according to the static indicators, and construct a fault-tolerant consistency constraint condition according to the deviation vector and the proportion of the number of deviation indicators, where strict consistency means that the states need to be exactly the same, and fault-tolerant consistency means that the state deviation needs to be less than or equal to the deviation threshold; based on the strict consistency constraint condition and the fault-tolerant consistency constraint condition, retrieve the obstetric care record data stored on the blockchain through the network; count the abnormal care trigger frequency ratio of the obstetric care record data, and set it as the abnormal care probability.

[0030] Specifically, strict consistency constraints are constructed based on static indicators. These static indicators are usually fixed and immutable, such as basic information like the age, gender, and pregnancy period of the care recipient. The strict consistency constraints require that the state to be predicted be exactly the same as the state described by these static indicators. Fault-tolerant consistency constraints are constructed based on the deviation vector and the proportion of the number of deviation indicators. The deviation vector reflects the deviation direction and degree of the dynamic indicators relative to the normal range, while the proportion of the number of deviation indicators refers to the proportion of the dynamic indicators that deviate from the normal range among all dynamic indicators. The fault-tolerant consistency constraints allow for a certain deviation in the state, but the deviation must be less than or equal to the preset deviation threshold. This means that as long as the deviation degree and the proportion of the number of deviations of the dynamic indicators are within the acceptable range, we consider the state to be consistent. After constructing the constraints, these conditions are used to retrieve the obstetric care record data stored on the blockchain through the network. Blockchain technology ensures the authenticity and immutability of the data, ensuring the accuracy of the data. Through the retrieval, obstetric care record data that matches the state to be predicted can be obtained. After obtaining the matching obstetric care record data, the trigger frequency of abnormal care events in these records is counted. Specifically, the ratio of the number of records that trigger abnormal care events to the total number of records is calculated, and this ratio is the abnormal care trigger frequency ratio. Finally, the abnormal care trigger frequency ratio is set as the abnormal care probability, and this probability value reflects the likelihood of an abnormal care event occurring under the given conditions. By comparing this probability value with the preset threshold, it can be determined whether preventive measures need to be taken or further examinations are required. In summary, by constructing strict consistency constraints and fault-tolerant consistency constraints, retrieving the obstetric care record data stored on the blockchain through the network, and counting the abnormal care trigger frequency ratio, the probability of abnormal care can be effectively predicted, providing strong support for the decision-making of obstetric care.

[0031] For a more intuitive explanation, the following example is given: Here, the static indicators are set as "pregnant woman's age = 30 years old" and "gestation period = 32 weeks". The strict consistency constraint requires that the age of the pregnant woman to be predicted must be 30 years old and the gestation period must be 32 weeks, and these two conditions must be met simultaneously. Suppose there are 5 dynamic indicators (such as blood pressure, heart rate, blood sugar, urine protein, fetal heart rate), and each indicator has a normal range. The deviation vector refers to the degree of deviation of the actual monitored values of these indicators from the normal range, which can be expressed in numerical values or percentages. Suppose the number of dynamic indicators deviating from the normal range does not exceed 2 (i.e., not exceeding 40% of the total number of indicators). The fault-tolerant consistency constraint allows the degree of deviation of the dynamic indicators to be within a certain range (such as ±10%), and the proportion of the number of indicators deviating from the normal range does not exceed 40%. After setting the requirements, a search is performed. Suppose there is a blockchain database containing 1000 obstetric care records, and each record contains the basic information of the pregnant woman, the monitored values of the dynamic indicators, and whether an abnormal care event is triggered. The obstetric care records that meet these conditions are retrieved according to the strict consistency constraint and the fault-tolerant consistency constraint. Suppose 200 obstetric care records that meet the conditions are retrieved, and among these 200 records, 30 have triggered abnormal care events. Then, the abnormal care trigger frequency ratio = the number of records triggering abnormal care events / the total number of records = 30 / 200 = 15%. Finally, the abnormal care probability is set to 15% according to the calculation result. This means that under the given conditions, there is a 15% possibility of an abnormal care event occurring.

[0032] In a preferred embodiment, based on the strict consistency constraint and the fault-tolerant consistency constraint, network search is performed on the obstetric care record data stored on the blockchain, including:

[0033] Based on the strict consistency constraint and the fault-tolerant consistency constraint, network search is performed on the first-level obstetric care record data stored on the blockchain; when the data volume of the first-level obstetric care record data is less than or equal to the data volume threshold, the first-level deviation vector and the proportion of the number of first-level deviation indicators are calculated, where the data volume threshold is greater than or equal to 5000 records; the fault-tolerant consistency constraint is updated according to the first-level deviation vector and the proportion of the number of first-level deviation indicators to obtain the first-level fault-tolerant consistency constraint, and combined with the strict consistency constraint, network search is performed on the second-level obstetric care record data stored on the blockchain; until the data volume of the first-level obstetric care record data to the N-level obstetric care record data is greater than or equal to the data volume threshold, or N is equal to 5, the first-level obstetric care record data to the N-level obstetric care record data are added to the obstetric care record data, where N is an integer.

[0034] Furthermore, during the process of predicting the probability of abnormal nursing, based on the strict consistency constraint conditions and the fault-tolerant consistency constraint conditions, the network retrieves the obstetric nursing record data stored on the blockchain. To avoid inaccurate prediction results caused by insufficient data volume, a hierarchical retrieval strategy is adopted. First, according to the strict consistency constraint conditions and the initially set fault-tolerant consistency constraint conditions, the network retrieves the first-level obstetric nursing record data stored on the blockchain, which are the records that best match the to-be-predicted status under the strict consistency and fault-tolerant consistency conditions. Then, it is judged whether the data volume of the retrieved first-level obstetric nursing record data is greater than or equal to the preset data volume threshold (for example, 5000 records). If the data volume is sufficient, these data are directly used as the final obstetric nursing record data for subsequent analysis. If the data volume of the first-level obstetric nursing record data is insufficient, the first-level deviation vector and the proportion of the number of first-level deviation indicators are calculated, and these two indicators reflect the deviation degree of the retrieved data from the to-be-predicted status in dynamic indicators. Then, according to this deviation information, the fault-tolerant consistency constraint conditions are updated to obtain the first-level fault-tolerant consistency constraint conditions to expand the retrieval scope. Subsequently, in combination with the updated strict consistency constraint conditions and the first-level fault-tolerant consistency constraint conditions, the network retrieves the second-level obstetric nursing record data stored on the blockchain. If the data volume of the second-level data is still insufficient, repeat the above steps, continue to calculate the deviation vector and the proportion of the number of deviation indicators, update the fault-tolerant consistency constraint conditions, and retrieve the next-level data. The hierarchical retrieval is continuously carried out until the data volume of the obstetric nursing record data at a certain level (for example, the Nth level) is greater than or equal to the data volume threshold, or when N is equal to 5 (that is, five levels of data have been retrieved, but the data volume threshold has not been reached), the retrieval process is terminated. Finally, all the obstetric nursing record data from the first level to the Nth level are integrated together as the final obstetric nursing record data for subsequent analysis.

[0035] In a preferred embodiment, the statistical abnormal nursing trigger frequency ratio of the obstetric nursing record data, set as the abnormal nursing probability, includes:

[0036] Calculating the deviation distance between the Nth-level deviation vector and the proportion of the number of Nth-level deviation indicators of the Nth-level obstetric nursing record data and the deviation vector and the proportion of the number of deviation indicators, set as the Nth-level distribution distance; until calculating the deviation distance between the Nth-level deviation vector and the proportion of the number of Nth-level deviation indicators of the first-level obstetric nursing record data and the deviation vector and the proportion of the number of deviation indicators, set as the first-level distribution distance; calculating the total distribution distance from the first-level distribution distance to the Nth-level distribution distance; calculating the ratio of the first-level distribution distance to the total distribution distance, and using 1 minus the ratio to generate the first-level distribution weight until obtaining the Nth-level distribution weight; weighted fusing the abnormal nursing trigger frequency ratios of each level according to the first-level distribution weight until the Nth-level distribution weight to obtain the abnormal nursing probability.

[0037] Specifically, in the process of counting the abnormal care trigger frequency ratio to set the abnormal care probability, a method of weighted fusion of multi-level obstetric care record data is adopted. This method considers the deviation distance (i.e., distribution distance) between different levels of data and the initial conditions in terms of the deviation vector and the proportion of the number of deviation indicators, and assigns fusion weights according to the proximity of the distribution distance. First, calculate the deviation vector and the proportion of the number of deviation indicators for the obstetric care record data at each level (from level 1 to level N), compare them with the initial deviation vector and the proportion of the number of deviation indicators, and calculate the deviation distance, which is called the distribution distance. The distribution distance reflects the degree of deviation of the data at this level from the initial conditions in terms of dynamic indicators. Then, accumulate the distribution distances from level 1 to level N to obtain the total distribution distance, which represents the overall degree of deviation of all levels of data from the initial conditions in terms of dynamic indicators. Next, calculate the ratio of the distribution distance of each level to the total distribution distance, and use 1 minus this ratio to generate the distribution weight for this level. The purpose of this step is to assign fusion weights according to the proximity of the distribution distance. The farther the distribution distance of a level is, the smaller the weight of its data in the fusion process. Finally, use the calculated distribution weights at each level to weighted-fuse the abnormal care trigger frequency ratio at each level. Specifically, multiply the abnormal care trigger frequency ratio at each level by the corresponding distribution weight, and then add up all the weighted frequency ratios to obtain the final abnormal care probability. In this process, the distribution distance plays a crucial role. It is not only an important indicator for measuring the degree of deviation of each level of data from the initial conditions, but also an important basis for assigning fusion weights. By calculating the distribution distance, it is possible to identify which levels of data are closer to the initial conditions, and thus give higher weights in the fusion process; while which levels of data deviate farther from the initial conditions, and thus give lower weights in the fusion process. This weighted fusion method based on the distribution distance can more accurately reflect the contribution degree of each level of data to the final abnormal care probability, thereby improving the accuracy and reliability of the prediction. To sum up, by calculating the distribution distances at each level, the total distribution distance, and the distribution weights at each level, and weighted-fusing the abnormal care trigger frequency ratios at each level, a more accurate abnormal care probability can be obtained. This method not only considers the quantity and quality of the data, but also considers the degree of deviation of the data from the initial conditions, and is a more comprehensive and accurate prediction method.

[0038] For a clearer understanding, the following is an example: Suppose there are three levels of obstetric care record data (level 1, level 2, level 3), and the deviation vector and the proportion of the number of deviation indicators for each level of data have been calculated. For simplicity, as shown in Table 1, assume that the deviation vector and the proportion of the number of deviation indicators can be represented by a comprehensive indicator called "deviation degree".

[0039] Table 1: Comparison table of deviation degree and distribution distance of three-level obstetric care record data

[0040]

[0041] Add up the distribution distances at all levels to obtain the total distribution distance. Total distribution distance = 0.05 (first level) + 0.05 (second level) + 0.2 (third level) = 0.3. Calculate the distribution weights at all levels based on the ratio of the distribution distance at each level to the total distribution distance, and use 1 minus the ratio to generate the weights, as shown in Table 2 (here, for simplicity, the weight results of 1 minus the ratio are directly given):

[0042] Table 2: Distribution Weight Table of Tertiary Obstetric Nursing Record Data with Simplified Calculation

[0043]

[0044] Note: In practical applications, since the total distribution distance is fixed, the sum of the distribution weights at all levels should be 1. The calculation method here is for simplified explanation. In practice, normalization should be performed according to the ratio of the distribution distance to the total to ensure that the sum of the weights is 1. As shown in Table 3, the correct weight calculation method is as follows:

[0045] Table 3: Distribution Weight Table of Tertiary Obstetric Nursing Record Data with Correct Calculation

[0046]

[0047] Assume that the abnormal care trigger frequency ratios at all levels are: 0.1 for the first level, 0.15 for the second level, and 0.2 for the third level. Perform weighted fusion based on the distribution weights at all levels: Abnormal care probability = 0.25 (first-level weight) × 0.1 (first-level frequency ratio) + 0.25 (second-level weight) × 0.15 (second-level frequency ratio) + 0.5 (third-level weight) × 0.2 (third-level frequency ratio) = 0.025 + 0.0375 + 0.1 = 0.1625. Therefore, the final abnormal care probability is 0.1625.

[0048] Please note that the data in practical applications will vary according to specific circumstances. In addition, when calculating the distribution weights, it should be ensured that the sum of the weights at all levels is 1 to avoid inaccurate weight allocation.

[0049] In a preferred embodiment, calculate the N-level deviation vector and the proportion of the number of N-level deviation indicators of the N-level obstetric nursing record data, and the deviation distance from the deviation vector and the proportion of the number of deviation indicators, which is set as the N-level distribution distance, including:

[0050] Calculate a first Euclidean distance between the N-level deviation vector and the deviation vector, where the first Euclidean distance has a first weight identifier; calculate a first proportion deviation between the proportion of the N-level deviation index quantity and the proportion of the deviation index quantity, where the first proportion deviation has a second weight identifier; according to the first weight identifier and the second weight identifier, perform weighted summation on the first Euclidean distance and the first proportion deviation to obtain the N-level distribution distance.

[0051] Exemplarily, when processing obstetric care record data, in order to evaluate the deviation degree between data at different levels (set as N levels) and the overall data, the "N-level distribution distance" can be calculated. This distance is a comprehensive index that combines the deviation vector of the N-level data and the deviation between the proportion of the deviation index quantity and the corresponding value of the overall data. First, it is necessary to obtain the N-level obstetric care record data, which contains the values of multiple care indicators. At the same time, it is also necessary to obtain the overall obstetric care record data as a comparison benchmark. For the N-level data and the overall data, their deviation vectors are calculated respectively. The deviation vector refers to the vector composed of the difference between each indicator value and its average value (or a certain benchmark value). Here, the N-level deviation vector specifically refers to the deviation vector of the N-level data, while the deviation vector refers to the deviation vector of the overall data. It is also necessary to calculate the proportion of the number of deviation indicators (i.e., those indicators with significant deviation from the average value) in the N-level data and the overall data. Next, calculate the first Euclidean distance between the N-level deviation vector and the deviation vector. The Euclidean distance is a commonly used distance metric method for measuring the straight-line distance between two vectors. This first Euclidean distance reflects the deviation degree of the N-level data from the overall data in terms of the deviation vector. To emphasize the importance of this distance, a first weight identifier is assigned to it, which indicates the weight of this distance when calculating the N-level distribution distance. The proportion of the deviation index quantity refers to the proportion of the number of deviation indicators in the total number of indicators. It is also necessary to calculate the first proportion deviation between the proportion of the N-level deviation index quantity and the proportion of the deviation index quantity. This deviation reflects the deviation degree of the N-level data from the overall data in terms of the proportion of the deviation index quantity. Similarly, to emphasize the importance of this deviation, a second weight identifier is given to it. Finally, according to the first weight identifier and the second weight identifier, perform weighted summation on the first Euclidean distance and the first proportion deviation to obtain the N-level distribution distance. This N-level distribution distance is a comprehensive index that simultaneously considers the deviation degrees of the N-level data from the overall data in terms of the deviation vector and the proportion of the deviation index quantity, and performs weighted processing according to the importance (i.e., weight) of these two deviations. Through the above process, a quantitative index - the N-level distribution distance - can be obtained to evaluate the deviation degree between the N-level obstetric care record data and the overall data. This index is of great significance for understanding the differences between data at different levels, optimizing care records, and improving the quality of care.

[0052] In a preferred embodiment, strict consistency constraint conditions are constructed according to the static indicators, and fault-tolerant consistency constraint conditions are constructed according to the deviation vector and the proportion of the number of deviation indicators, including:

[0053] Based on the static indicators, strict consistency reference parameters are constructed. When the input indicators are exactly the same as the strict consistency reference parameters, it is regarded as meeting the strict consistency constraint conditions; based on the deviation vector and the vector distance threshold, the first fault-tolerant consistency reference parameter is constructed, and based on the proportion of the number of deviation indicators and the proportion deviation threshold, the second fault-tolerant consistency reference parameter is constructed; when the deviation distance between the input deviation vector and the first fault-tolerant consistency reference parameter is less than or equal to the vector distance threshold, and the proportion deviation between the input proportion of the number of deviation indicators and the second fault-tolerant consistency reference parameter is less than or equal to the proportion deviation threshold, it is regarded as meeting the fault-tolerant consistency constraint conditions.

[0054] Specifically, when dealing with a set of static metrics, in order to ensure the accuracy and consistency of data, two types of constraint conditions can be constructed: strict consistency constraint conditions and fault-tolerant consistency constraint conditions. Identify the static metrics that need to be evaluated, as these metrics are the benchmarks for data accuracy and consistency. Based on these static metrics, construct a set of strictly consistent benchmark parameters, which includes the ideal values or standard values of each static metric. When the input metric data is exactly the same as the strictly consistent benchmark parameters, it is considered that the data meets the strict consistency constraint conditions. This means that the data is completely consistent with the benchmark parameters in all aspects without any deviation. For the input data, first calculate its deviation vector, which is a vector composed of the differences between each metric and the benchmark value. At the same time, also calculate the proportion of the number of deviated metrics, that is, the proportion of the number of metrics that deviate from the benchmark value to the total number of metrics. To construct the fault-tolerant consistency constraint conditions, two thresholds need to be set: the vector distance threshold and the proportion deviation threshold. These two thresholds are used to measure the tolerance ranges of the deviation vector and the proportion of the number of deviated metrics respectively. Based on the deviation vector and the vector distance threshold, construct the first fault-tolerant consistency benchmark parameter. This parameter is a benchmark that allows the deviation vector to fluctuate within a certain range. Similarly, based on the proportion of the number of deviated metrics and the proportion deviation threshold, construct the second fault-tolerant consistency benchmark parameter. This parameter is a benchmark that allows the proportion of the number of deviated metrics to fluctuate within a certain range. When the deviation distance between the input deviation vector and the first fault-tolerant consistency benchmark parameter is less than or equal to the vector distance threshold, and the proportion deviation between the input proportion of the number of deviated metrics and the second fault-tolerant consistency benchmark parameter is less than or equal to the proportion deviation threshold, it is considered that the data meets the fault-tolerant consistency constraint conditions. This means that although the data has a certain deviation from the benchmark parameters, the deviation is within an acceptable range, so it is still considered consistent. By constructing these two constraint conditions, a more detailed and comprehensive evaluation of data consistency and accuracy can be carried out. The strict consistency constraint conditions ensure the complete consistency of the data, while the fault-tolerant consistency constraint conditions allow deviations within a certain range, thus improving the flexibility and practicality of the evaluation.

[0055] The real-time obstetric care optimization method based on big data provided by the embodiments of the present invention has at least the following technical effects:

[0056] 1. By statistically analyzing the dynamic metrics in the abnormal care logs, distinguishing the central dynamic metrics and the marginal dynamic metrics, and calculating their deviation vectors and the proportion of the number of deviated metrics. Combining with the static metrics, it is possible to accurately predict the probability of abnormal care that meets these conditions, thus timely reminding the nursing side to optimize the care. This big data-based analysis method improves the recognition accuracy and efficiency of abnormal care risks.

[0057] 2. A multi-level retrieval strategy is adopted. Starting from the first-level obstetric care record data, it delves deeper level by level according to conditions such as the data volume threshold and distribution distance until sufficient data volume is obtained or the highest level (Nth level) is reached. By calculating the deviation vector and the proportion of the number of deviation indicators for each level of data, and weighted-fusing the abnormal care trigger frequency ratios of each level, the abnormal care probability is finally obtained. This method of multi-level retrieval and fusion effectively improves the accuracy and reliability of prediction.

[0058] 3. When predicting the abnormal care probability, strict consistency constraint conditions and fault-tolerant consistency constraint conditions are constructed. The strict consistency constraint conditions ensure that the input indicators are exactly the same as the benchmark parameters, while the fault-tolerant consistency constraint conditions allow the input indicators to deviate from the benchmark parameters within a certain range. The combined use of these two constraint conditions not only ensures the strict consistency of the data but also improves the fault-tolerance ability of the system, making the prediction results more in line with the actual situation.

[0059] Embodiment 2:

[0060] As Figure 2 shown, based on the same inventive concept as the method for real-time obstetric care optimization based on big data provided in Embodiment 1, the present invention embodiment also provides a system for real-time obstetric care optimization based on big data, including:

[0061] An information retrieval module 11, configured to use the static indicators of the obstetric care object as constants and the dynamic indicators as variables to retrieve multiple abnormal care logs, where any one of the abnormal care logs includes a set of abnormal dynamic indicators.

[0062] A statistical screening module 12, configured to, based on the set of abnormal dynamic indicators, statistically screen the central dynamic indicators whose number of occurrence logs in the multiple abnormal care logs is greater than or equal to the log number threshold, and the marginal dynamic indicators whose number of occurrence logs is less than the log number threshold.

[0063] A deviation statistics module 13, configured to statistically calculate the deviation vector of the monitoring values of the central dynamic indicators of the obstetric care object and the proportion of the number of deviation indicators of the monitoring values of the marginal dynamic indicators.

[0064] An abnormal prediction module 14, configured to predict the abnormal care probability that simultaneously meets the static indicators, the deviation vector, and the proportion of the number of deviation indicators.

[0065] A care optimization module 15, configured to, when the abnormal care probability is greater than or equal to the abnormal probability threshold, remind the nursing end to perform care optimization.

[0066] Furthermore, the statistical screening module 12 is further configured to perform the following steps:

[0067] Extract the first abnormal dynamic index set of the multiple abnormal nursing logs, where the first abnormal dynamic index set is statistically obtained in a preset length time zone before the care object shows abnormalities; count the first abnormal dynamic index deviation frequency of the first abnormal dynamic index in the preset length time zone in the first abnormal dynamic index set; when the first abnormal dynamic index deviation frequency is greater than or equal to the deviation frequency threshold, consider the number of logs with the first abnormal dynamic index appearing to increase by one; traverse the dynamic indexes, and based on the multiple abnormal nursing logs, perform statistics to obtain the central dynamic indexes with the number of logs appearing greater than or equal to the log number threshold, and the marginal dynamic indexes with the number of logs appearing less than the log number threshold.

[0068] Furthermore, the abnormal prediction module 14 is further configured to perform the following steps:

[0069] According to the static indexes, construct a strict consistency constraint condition, and according to the deviation vector and the deviation index quantity ratio, construct a fault-tolerant consistency constraint condition, where strict consistency means that the states need to be exactly the same, and fault-tolerant consistency means that the state deviation needs to be less than or equal to the deviation threshold; based on the strict consistency constraint condition and the fault-tolerant consistency constraint condition, retrieve the obstetric nursing record data stored on the blockchain through the network; count the abnormal nursing trigger frequency ratio of the obstetric nursing record data, and set it as the abnormal nursing probability.

[0070] Furthermore, the abnormal prediction module 14 is further configured to perform the following steps:

[0071] Based on the strict consistency constraint condition and the fault-tolerant consistency constraint condition, retrieve the first-level obstetric nursing record data stored on the blockchain through the network; when the data volume of the first-level obstetric nursing record data is less than or equal to the data volume threshold, calculate the first-level deviation vector and the first-level deviation index quantity ratio, where the data volume threshold is greater than or equal to 5000 records; update the fault-tolerant consistency constraint condition according to the first-level deviation vector and the first-level deviation index quantity ratio to obtain the first-level fault-tolerant consistency constraint condition, and in combination with the strict consistency constraint condition, retrieve the second-level obstetric nursing record data stored on the blockchain through the network; until the data volume of the first-level obstetric nursing record data until the N-level obstetric nursing record data is greater than or equal to the data volume threshold, or N is equal to 5, add the first-level obstetric nursing record data until the N-level obstetric nursing record data to the obstetric nursing record data, and N is an integer.

[0072] Furthermore, the abnormal prediction module 14 is further configured to perform the following steps:

[0073] Calculate the N-level deviation vector of the N-level obstetric care record data and the proportion of the number of N-level deviation indicators, and the deviation distance from the deviation vector and the proportion of the number of deviation indicators, which is set as the N-level distribution distance; until calculating the N-level deviation vector of the first-level obstetric care record data and the proportion of the number of N-level deviation indicators, and the deviation distance from the deviation vector and the proportion of the number of deviation indicators, which is set as the first-level distribution distance; calculate the total distribution distance from the first-level distribution distance to the N-level distribution distance; calculate the ratio of the first-level distribution distance to the total distribution distance, and use 1 minus the ratio to generate the first-level distribution weight until obtaining the N-level distribution weight; weighted fuse the abnormal care trigger frequency ratios of each level according to the first-level distribution weight until the N-level distribution weight to obtain the abnormal care probability.

[0074] Furthermore, the abnormal prediction module 14 is further configured to perform the following steps:

[0075] Calculate the first Euclidean distance between the N-level deviation vector and the deviation vector, where the first Euclidean distance has a first weight identifier; calculate the first proportion deviation between the proportion of the number of N-level deviation indicators and the proportion of the number of deviation indicators, where the first proportion deviation has a second weight identifier; according to the first weight identifier and the second weight identifier, perform weighted summation on the first Euclidean distance and the first proportion deviation to obtain the N-level distribution distance.

[0076] Furthermore, the abnormal prediction module 14 is further configured to perform the following steps:

[0077] Based on the static indicators, construct a strictly consistent benchmark parameter. When the input indicator is exactly the same as the strictly consistent benchmark parameter, it is regarded as meeting the strictly consistent constraint condition; based on the deviation vector and the vector distance threshold, construct a first fault-tolerant consistent benchmark parameter, and based on the proportion of the number of deviation indicators and the proportion deviation threshold, construct a second fault-tolerant consistent benchmark parameter; when the deviation distance between the input deviation vector and the first fault-tolerant consistent benchmark parameter is less than or equal to the vector distance threshold, and the proportion deviation between the input proportion of the number of deviation indicators and the second fault-tolerant consistent benchmark parameter is less than or equal to the proportion deviation threshold, it is regarded as meeting the fault-tolerant consistent constraint condition.

[0078] Through the foregoing detailed description of a real-time obstetric care optimization method based on big data in this specification, those skilled in the art can clearly know a real-time obstetric care optimization system based on big data in this embodiment. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0079] The foregoing description of the disclosed embodiments enables those skilled in the art to practice or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An obstetric real-time care optimization method based on big data, characterized in that, Including: Taking the static indicators of the obstetric care object as constants and the dynamic indicators as variables, retrieving multiple abnormal care logs. Any one of the abnormal care logs includes a set of abnormal dynamic indicators. The static indicators represent indicators that are relatively stable in the short term and are not likely to change significantly. The dynamic indicators represent indicators that will change with time, environment, or treatment intervention; Based on the set of abnormal dynamic indicators, counting the central dynamic indicators for which the number of occurrence logs of the dynamic indicators in the multiple abnormal care logs is greater than or equal to the log number threshold, and the marginal dynamic indicators for which the number of occurrence logs is less than the log number threshold; Counting the deviation vector of the monitoring values of the central dynamic indicators of the obstetric care object and the proportion of the number of deviation indicators of the monitoring values of the marginal dynamic indicators; Predicting the abnormal care probability that simultaneously meets the static indicators, the deviation vector, and the proportion of the number of deviation indicators; When the abnormal care probability is greater than or equal to the abnormal probability threshold, reminding the nursing end to optimize the care; Among them, predicting the abnormal care probability that simultaneously meets the static indicators, the deviation vector, and the proportion of the number of deviation indicators includes: Constructing a strict consistency constraint condition according to the static indicators, and constructing a fault-tolerant consistency constraint condition according to the deviation vector and the proportion of the number of deviation indicators. Among them, strict consistency means that the states need to be exactly the same, and fault-tolerant consistency means that the state deviation needs to be less than or equal to the deviation threshold; Based on the strict consistency constraint condition and the fault-tolerant consistency constraint condition, retrieving the obstetric care record data stored on the blockchain through networking; Counting the abnormal care trigger frequency ratio of the obstetric care record data and setting it as the abnormal care probability; Among them, retrieving the obstetric care record data stored on the blockchain through networking based on the strict consistency constraint condition and the fault-tolerant consistency constraint condition includes: Based on the strict consistency constraint condition and the fault-tolerant consistency constraint condition, retrieving the first-level obstetric care record data stored on the blockchain through networking; When the data volume of the first-level obstetric care record data is less than or equal to the data volume threshold (where the data volume threshold is equal to 5000 records), calculating the first-level deviation vector and the proportion of the number of first-level deviation indicators; Updating the fault-tolerant consistency constraint condition according to the first-level deviation vector and the proportion of the number of first-level deviation indicators to obtain the first-level fault-tolerant consistency constraint condition, and combining the strict consistency constraint condition, retrieving the second-level obstetric care record data stored on the blockchain through networking; When the data volume of the first-level obstetric care record data to the N-level obstetric care record data is greater than or equal to the data volume threshold, or N is equal to 5, adding the first-level obstetric care record data to the N-level obstetric care record data to the obstetric care record data, where N is an integer; Among them, counting the abnormal care trigger frequency ratio of the obstetric care record data and setting it as the abnormal care probability includes: Calculating the deviation distance between the first-level deviation vector and the proportion of the number of first-level deviation indicators of the first-level obstetric care record data and the deviation vector and the proportion of the number of deviation indicators, and setting it as the first-level distribution distance; Until calculating the N-level deviation vector of the N-level obstetric care record data and the proportion of the number of N-level deviation indicators, and the deviation distance from the deviation vector and the proportion of the number of deviation indicators, which is set as the N-level distribution distance; Calculate the total distribution distance from the first-level distribution distance to the N-level distribution distance; Calculate the ratio of the first-level distribution distance to the total distribution distance, and use 1 minus the ratio to generate the first-level distribution weight until obtaining the N-level distribution weight; According to the first-level distribution weight until the N-level distribution weight, weighted fusion of the abnormal care trigger frequency ratio of each level to obtain the abnormal care probability.

2. The method according to claim 1, wherein Based on the abnormal dynamic index set, count the central dynamic indexes of the dynamic indexes whose occurrence log count in the multiple abnormal care logs is greater than or equal to the log count threshold, and the marginal dynamic indexes whose occurrence log count is less than the log count threshold, including: Extract the first abnormal dynamic index set of the first abnormal care log of the multiple abnormal care logs, where the first abnormal dynamic index set is counted in a preset length time zone before the care object shows abnormality; Count the first abnormal dynamic index deviation frequency of the first abnormal dynamic index in the first abnormal dynamic index set in the preset length time zone; When the first abnormal dynamic index deviation frequency is greater than or equal to the deviation frequency threshold, consider the occurrence log count of the first abnormal dynamic index to be incremented by one; Traverse the dynamic indexes and count based on the multiple abnormal care logs to obtain the central dynamic indexes whose occurrence log count is greater than or equal to the log count threshold, and the marginal dynamic indexes whose occurrence log count is less than the log count threshold.

3. The method according to claim 1, characterized in that, Calculate the deviation distance between the N-level deviation vector of the N-level obstetric care record data and the proportion of the number of N-level deviation indicators, and the deviation vector and the proportion of the number of deviation indicators, which is set as the N-level distribution distance, including: Calculate the first Euclidean distance between the N-level deviation vector and the deviation vector, where the first Euclidean distance has a first weight identifier; Calculate the first proportion deviation between the proportion of the number of N-level deviation indicators and the proportion of the number of deviation indicators, where the first proportion deviation has a second weight identifier; According to the first weight identifier and the second weight identifier, perform weighted summation on the first Euclidean distance and the first proportion deviation to obtain the N-level distribution distance.

4. The method according to claim 1, wherein According to the static indicators, construct strict consistency constraint conditions, and according to the deviation vector and the proportion of the number of deviation indicators, construct fault-tolerant consistency constraint conditions, including: Based on the static indicators, construct strict consistency reference parameters. When the input indicators are exactly the same as the strict consistency reference parameters, it is considered to meet the strict consistency constraint conditions; Based on the deviation vector and the vector distance threshold, construct the first fault-tolerant consistency reference parameter, and based on the proportion of the number of deviation indicators and the proportion deviation threshold, construct the second fault-tolerant consistency reference parameter; When the deviation distance between the input deviation vector and the first fault tolerance consistent reference parameter is less than or equal to the vector distance threshold, and the proportion deviation between the proportion of the input deviation index quantity and the second fault tolerance consistent reference parameter is less than or equal to the proportion deviation threshold, it is considered that the fault tolerance consistent constraint condition is satisfied.

5. An obstetric real-time nursing optimization system based on big data, characterized in that, For implementing a real-time obstetric care optimization method based on big data according to any one of claims 1-4, the system includes: An information retrieval module, configured to retrieve multiple abnormal care logs with static indicators of an obstetric care object as constants and dynamic indicators as variables, wherein any one of the abnormal care logs includes a set of abnormal dynamic indicators; A statistical screening module, configured to, based on the set of abnormal dynamic indicators, statistically screen central dynamic indicators whose occurrence log count in the multiple abnormal care logs is greater than or equal to the log count threshold, and marginal dynamic indicators whose occurrence log count is less than the log count threshold; A deviation statistics module, configured to statistically calculate the deviation vector of the monitored value of the central dynamic indicator of the obstetric care object and the proportion of the deviation index quantity of the monitored value of the marginal dynamic indicator; An abnormal prediction module, configured to predict the abnormal care probability that simultaneously satisfies the static indicator, the deviation vector, and the proportion of the deviation index quantity; A care optimization module, configured to remind the care side to perform care optimization when the abnormal care probability is greater than or equal to the abnormal probability threshold.

Images