Gestational diabetes mellitus dynamic risk assessment method and system

Abstract

The invention provides a gestational diabetes mellitus dynamic risk assessment method and a gestational diabetes mellitus dynamic risk assessment system, and relates to the technical field of gestational diabetes mellitus risk assessment. The collaborative fluctuation intensity of the three core characteristics is accurately quantified by using a matrix norm, a two-factor decision engine is generated by combining postprandial blood sugar daily mean value change direction marks, a long-term metabolism trend under a gestational week differentiation weight is quantified by using a risk coefficient, and a linkage factor is fused with multi-characteristic collaborative fluctuation and a blood sugar short-term change direction. A blood glucose threshold self-adaptive adjustment algorithm is driven in real time, and a traditional static threshold system is thoroughly broken through; an acute event hard interception network is synchronously established, risk assessment result forced coverage and minute-level monitoring activation are triggered through a linkage factor threshold value, and a mechanical response barrier for acute metabolism decompensation is formed; and finally, dynamic risk assessment of the gestational diabetes mellitus is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of gestational diabetes risk assessment, and particularly relates to a gestational diabetes dynamic risk assessment method and system. BACKGROUND

[0002] Gestational diabetes mellitus (GDM) is the most common metabolic disease during pregnancy, with a global incidence of 10-15%. It significantly increases the risk of maternal and infant complications, including macrosomia, preterm birth, and long-term diabetes in offspring. Current clinical practice relies on fixed blood glucose thresholds for risk assessment, such as a static standard of fasting blood glucose not exceeding 5.1 mmol / L or postprandial blood glucose not exceeding 7.8 mmol / L. However, this model has fundamental flaws. First, static thresholds cannot adapt to the dynamic evolution of metabolism during pregnancy, especially the exacerbation of insulin resistance in the third trimester, which doubles the clinical risk of the same blood glucose value. Second, the existing assessment system only focuses on absolute blood glucose values, completely ignoring the synergistic effects of multiple metabolic characteristics such as fasting glucose variability, postprandial glucose decline rate, and body weight growth trend. More seriously, traditional methods have a severe lag in response to acute metabolic events, failing to capture dangerous signals such as postprandial blood glucose spikes within 24 hours, resulting in missed golden intervention windows for conditions such as ketoacidosis. In addition, the phenomenon of sudden weight gain is not distinguished from pregnancy edema interference, greatly weakening the reliability of the assessment.

[0003] Therefore, it is necessary to provide a gestational diabetes dynamic risk assessment method and system to solve the above technical problems. SUMMARY

[0004] To solve the above technical problems, the present application provides a gestational diabetes dynamic risk assessment method and system, which achieves the beneficial effect of accurately and effectively assessing the risk of gestational diabetes. The present application provides a gestational diabetes dynamic risk assessment method, comprising: S1: obtaining a gestational monitoring data set of a pregnant woman during pregnancy, the gestational monitoring data set comprising current gestational age, fasting blood glucose value record, postprandial blood glucose value record, body weight record, and daily carbohydrate intake estimate; S2: determining the gestational metabolic stage of the current gestational age based on the gestational metabolic stage divided by gestational age, and calculating the fasting blood glucose coefficient of variation, postprandial blood glucose decline slope, and body weight growth deviation based on the gestational monitoring data set and in combination with the gestational metabolic stage of the current gestational age; S3: constructing a dynamic linkage matrix based on the fasting blood glucose coefficient of variation, postprandial blood glucose decline slope, and body weight growth deviation, and calculating the risk coefficient and linkage factor in combination with the gestational metabolic stage of the current gestational age; S4: Obtain the fasting blood glucose benchmark threshold and the postprandial blood glucose benchmark threshold of the current gestational age from the clinical guideline library, and adjust the fasting blood glucose benchmark threshold through the linkage factor, and adjust the postprandial blood glucose benchmark threshold through the linkage factor and the body weight gain deviation; S5: Generate a multi-level risk assessment result based on the adjusted fasting blood glucose benchmark threshold and the postprandial blood glucose benchmark threshold combined with the risk coefficient.

[0005] Preferably, in step S2, the calculation of the fasting blood glucose coefficient of variation includes the following steps: Obtain the fasting blood glucose value records of the last week from the current day from the pregnancy monitoring data set, and calculate the fasting blood glucose standard deviation and the fasting blood glucose mean value; Based on the pregnancy metabolic stage of the current gestational age, a preset coefficient of variation formula is used to process the fasting blood glucose standard deviation and the fasting blood glucose mean value to obtain the fasting blood glucose coefficient of variation.

[0006] Preferably, in step S2, the calculation of the postprandial blood glucose decline slope includes the following steps: Obtain the postprandial blood glucose value records of the last week from the current day from the pregnancy monitoring data set, and obtain multiple postprandial blood glucose curves for each day through the interpolation method; Identify the postprandial blood glucose peak value of each postprandial blood glucose curve, and determine the fasting blood glucose correction baseline according to the pregnancy metabolic stage of the current gestational age and the daily fasting blood glucose value record; Take the data segment of each postprandial blood glucose curve from the postprandial blood glucose peak point to the first time the blood glucose value is less than or equal to the correction baseline, use the least squares method to fit a linear regression equation, and take the absolute value of the regression slope as the blood glucose decline slope of a single meal, while calculating the fitting error; Determine the fusion weight of the blood glucose decline slope of a single meal based on the fitting error and perform weighted fusion to obtain the daily postprandial blood glucose decline slope, and obtain the postprandial blood glucose decline slope by weighted fusion of the daily postprandial blood glucose decline slope.

[0007] Preferably, in step S2, the calculation step of the body weight gain deviation includes the following steps: Obtain the body weight records of the last week from the current day from the pregnancy monitoring data set, and calculate the actual body weight growth rate in the last week from the current day; Obtain the body weight recommended growth rate in the clinical guideline according to the pregnancy metabolic stage of the current gestational age, and calculate the body weight gain deviation combined with the actual body weight growth rate.

[0008] Preferably, in step S3, the structure of the dynamic linkage matrix is: The first row of the matrix is the fasting blood glucose coefficient of variation, the postprandial blood glucose decline slope, and the body weight gain deviation of the current day; The second row of the matrix is empty, and the coefficient of variation of fasting blood glucose, the daily variation of the postprandial blood glucose decline slope, and the daily variation of the body weight growth deviation degree.

[0009] Preferably, in step S3, the fasting blood glucose variation coefficient, the postprandial blood glucose decline slope, and the body weight growth deviation degree are weighted and fused according to the metabolic stage of the current gestational age to obtain a risk coefficient.

[0010] Preferably, in step S3, the calculation step of the linkage factor includes: The Frobenius norm is calculated based on all elements of the dynamic linkage matrix to obtain the overall fluctuation intensity; The difference between the postprandial blood glucose mean value of all meals on the current day and the previous day is calculated, and a sign function is combined to obtain a postprandial blood glucose variation direction marker value, and the overall fluctuation intensity and the postprandial blood glucose variation direction marker value are weighted and fused to generate the linkage factor.

[0011] Preferably, in step S4, the adjustment step of the fasting blood glucose reference threshold and the postprandial blood glucose reference threshold includes: The fasting blood glucose reference adjustment amount is obtained by multiplying the gestational age, the linkage factor, and the preset fasting blood glucose reference adjustment coefficient, and the adjusted fasting blood glucose reference threshold is obtained by multiplying the fasting blood glucose reference threshold by 1 plus the fasting blood glucose adjustment coefficient; The postprandial blood glucose reference adjustment amount is obtained by multiplying the body weight growth deviation degree, the linkage factor, and the preset postprandial blood glucose reference adjustment coefficient, and the adjusted postprandial blood glucose reference threshold is obtained by multiplying the postprandial blood glucose reference threshold by 1 plus the postprandial blood glucose adjustment coefficient.

[0012] Preferably, step S5 further includes an emergency triggering mechanism based on the linkage factor and a preset emergency triggering threshold, including: When the linkage factor exceeds the preset emergency triggering threshold, the risk assessment result is forcibly covered as the highest risk, and the preset emergency response measure is started.

[0013] The application also provides a gestational diabetes mellitus dynamic risk assessment system applied to the gestational diabetes mellitus dynamic risk assessment method, including: A data acquisition module is configured to obtain a gestational monitoring data set of a pregnant woman, the gestational monitoring data set including a current gestational age, fasting blood glucose value records, postprandial blood glucose value records, body weight records, and daily carbohydrate intake estimates; A metabolic feature dynamic extraction module is configured to determine a gestational metabolic stage of the current gestational age based on a gestational metabolic stage divided by the gestational age, and calculate a fasting blood glucose variation coefficient, a postprandial blood glucose decline slope, and a body weight growth deviation degree based on the gestational monitoring data set and in combination with the gestational metabolic stage of the current gestational age; A synergy matrix construction module is configured to construct a dynamic linkage matrix based on the fasting blood glucose variation coefficient, postprandial blood glucose decline slope and body weight growth deviation degree, and calculate a risk coefficient and a linkage factor in combination with a pregnancy metabolic stage of the current gestational age; A blood glucose threshold dynamic adjustment module is configured to obtain fasting blood glucose reference thresholds and postprandial blood glucose reference thresholds of the current gestational age from a clinical guideline library, and adjust the fasting blood glucose reference thresholds through the linkage factor and adjust the postprandial blood glucose reference thresholds through the linkage factor and the body weight growth deviation degree; A multi-level risk assessment module is configured to generate a multi-level risk assessment result based on the adjusted fasting blood glucose reference thresholds and postprandial blood glucose reference thresholds in combination with the risk coefficient.

[0014] Compared with the related art, the method and system for dynamically assessing the risk of gestational diabetes mellitus provided by the application have the following beneficial effects: The application realizes a fundamental breakthrough in the paradigm of risk assessment of gestational diabetes mellitus by means of a multi-dimensional metabolic feature dynamic coupling and real-time response mechanism, constructs a dynamic linkage matrix based on the fasting blood glucose variation coefficient, postprandial blood glucose decline slope and body weight growth deviation degree, precisely quantifies the collaborative fluctuation intensity of the three core features by means of the matrix Frobenius norm, generates a two-factor decision engine in combination with the direction marker of the daily average value of postprandial blood glucose, quantifies the long-term metabolic trend under the differentiated weight of the risk coefficient and gestational age, fuses the multi-feature collaborative fluctuation and the direction of short-term blood glucose change by means of the linkage factor, and real-time drives the blood glucose threshold self-adaptive adjustment algorithm, thereby completely breaking through the traditional static threshold system; an acute event hard blocking network is simultaneously established, the risk assessment result is forcibly covered and minute-level monitoring is activated by means of the linkage factor threshold, thereby forming a mechanical response barrier to acute metabolic decompensation; a multi-level fault-tolerant architecture is integrated, thereby realizing three-way anti-interference protection of the data layer, feature layer and model layer by means of the postprandial blood glucose curve interpolation residual error control mechanism, body weight growth rate edema interference correction module and matrix rolling update protocol, and finally, realizing the dynamic risk assessment of gestational diabetes mellitus with real-time collaborative analysis capability and complex scene robustness. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 A flowchart of the method for dynamically assessing the risk of gestational diabetes mellitus according to the application; Figure 2 A module structure diagram of the system for dynamically assessing the risk of gestational diabetes mellitus according to the application. DETAILED DESCRIPTION

[0016] The application will be described in further detail below with reference to the drawings and embodiments. It is to be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application. In addition, it is to be understood that the drawings are not necessarily to scale and that, unless otherwise indicated, the drawings are merely intended to schematically depict the detailed description. Embodiments and features described herein can be combined with each other, as much as is not contradictory.

[0017] In addition, it is to be understood that the drawings are not necessarily to scale and that, unless otherwise indicated, the drawings are merely intended to schematically depict the detailed description. Embodiments and features described herein can be combined with each other, as much as is not contradictory.

[0018] Embodiment one A dynamic risk assessment method for gestational diabetes, in the specific implementation process, as shown in the figure, it shows a flow chart of a dynamic risk assessment method for gestational diabetes, including: Figure 1 Step S1: obtaining the gestational monitoring data set of pregnant women during pregnancy, the gestational monitoring data set including current gestational age, fasting blood glucose value record, postprandial blood glucose value record, body weight record and daily carbohydrate intake estimate value.

[0019] ​In the implementation process, in the data acquisition stage, the application constructs a pregnancy monitoring dataset through multi-modal integration and structured management, provides a standardized input source for dynamic risk assessment, and the system obtains five types of core data from the pregnant woman's daily monitoring terminal in real time, the current gestational age, automatically calculates and accurately to the day according to the last menstrual period or ultrasound calibration data; fasting blood glucose value record, the first measurement value after getting up in the morning is recorded in millimoles per liter; postprandial blood glucose value record, covering multiple measurement values within one to two hours after breakfast, lunch and dinner and marking the specific meal and measurement timestamp; body weight record, the measurement value in the state of empty stomach in the morning is recorded in kilograms; daily carbohydrate intake estimated value, based on the user's diet log, the preset food database is converted into grams of carbohydrates. All records are integrated according to the natural day, missing data triggers an automatic filling agreement, for example, if a single postprandial blood glucose is missing, the average value of the same meal for the previous three days is used instead, and if the whole day body weight is not recorded, it is marked as an invalid day but does not interrupt the weekly statistical process; in addition, the collected data also needs to be preprocessed for subsequent calculation and use, for example, an abnormal filtering rule is established, when the fasting blood glucose value is lower than 2.8 millimoles per liter or higher than 7.0 millimoles per liter, an automatic review reminder is triggered, when the single-day weight gain is greater than 0.5 kilograms, the associated edema evaluation mark needs to be confirmed by the user for interference factors, through the establishment of a metabolic panoramic view, the causal relationship between carbohydrate intake and blood glucose fluctuation is formed, and the limitation of isolated analysis of blood glucose value in traditional evaluation is broken; with the help of time sequence alignment, the timeliness of dynamic modeling is ensured, and the accuracy of daily change calculation is ensured; relying on multi-source verification and abnormal filtering mechanism, noise interference is reduced from the source, and a robust foundation is laid for subsequent feature extraction.

[0020] Step S2: determining the pregnancy metabolic stage of the current gestational age based on the gestational age division, calculating the fasting blood glucose coefficient of variation, the postprandial blood glucose decline slope and the body weight gain deviation degree based on the pregnancy monitoring dataset and combining the pregnancy metabolic stage of the current gestational age.

[0021] In the implementation process, through gestational age driven metabolic stage division and multi-dimensional feature fusion, the dynamic quantification of the metabolic state during pregnancy is realized, first of all, according to the International Federation of Gynecology and Obstetrics guidelines, the pregnancy period is divided into early pregnancy (1-12 weeks), mid-pregnancy (13-27 weeks) and late pregnancy (28-40 weeks), each stage corresponds to different metabolic feature calculation rules, based on this, three core features are calculated in parallel: fasting blood glucose coefficient of variation: reflects the stability of the basic blood glucose, in the third trimester of pregnancy, the fluctuation sensitivity needs to be amplified due to the aggravation of insulin resistance; postprandial blood glucose decline slope: represents the function of glucose tolerance, in the second trimester of pregnancy, a stricter decline rate standard is adopted; body weight gain deviation degree: evaluates the balance of nutrient metabolism, in the third trimester of pregnancy, the disturbance of fluid retention is mainly monitored.

[0022] Specifically, in step S2, the calculation of the fasting blood glucose coefficient of variation includes the following steps: Obtaining the fasting blood glucose value records of the last week from the current day in the pregnancy monitoring data set and calculating the fasting blood glucose standard deviation and fasting blood glucose mean value; Selecting a preset coefficient of variation formula based on the current gestational age of the pregnancy metabolic stage to process the fasting blood glucose standard deviation and fasting blood glucose mean value to obtain the fasting blood glucose coefficient of variation.

[0023] In the specific implementation process, the fasting blood glucose records of the last seven consecutive days are extracted from the pregnancy monitoring data set, the blood glucose standard deviation and arithmetic mean value in the period are calculated first, and then a preset formula is selected according to the metabolic stage to which the current gestational age belongs. For example, the classic coefficient of variation formula of standard deviation divided by mean value is directly used in the early pregnancy; and the correction formula is used in the late pregnancy, and the coefficient of variation is equal to the standard deviation multiplied by 1.2 and then divided by the mean value. The correction is derived from the physiological characteristics that the fluctuation of liver glucose output is significantly enhanced in the middle and late pregnancy. The calculation process automatically excludes the interference of extreme values. If the fasting blood glucose of a single day is lower than 2.8 or higher than 7.0 mmol / L, it is replaced by the mean value of the previous three days, so as to ensure that the coefficient of variation truly reflects the stability of the basic blood glucose rather than the measurement error.

[0024] Specifically, in step S2, the calculation of the postprandial blood glucose decline slope includes the following steps: Obtaining the postprandial blood glucose value records of the last week from the current day in the pregnancy monitoring data set, and obtaining multiple postprandial blood glucose curves of each day by interpolation method; Identifying the postprandial blood glucose peak value of each postprandial blood glucose curve, and determining the fasting blood glucose correction baseline according to the current gestational age of the pregnancy metabolic stage and the fasting blood glucose value record of each day; Cutting the data segment of each postprandial blood glucose curve from the postprandial blood glucose peak point to the point where the blood glucose value is less than or equal to the correction baseline for the first time, fitting a linear regression equation by least square method and taking the absolute value of the regression slope as the blood glucose decline slope of a single meal, and calculating the fitting error; Determining the fusion weight of the blood glucose decline slope of a single meal based on the fitting error and performing weighted fusion to obtain the postprandial blood glucose decline slope of each day, and obtaining the postprandial blood glucose decline slope by weighted fusion of the postprandial blood glucose decline slope of each day.

[0025] In the implementation process, for the postprandial blood glucose value of each meal in the last seven days, for example, first generate a continuous blood glucose curve by three times spline interpolation method, each curve needs to meet the accuracy constraint of passing all original data points and the residual sum of squares is not greater than 0.2. After identifying the postprandial blood glucose peak point, the fasting blood glucose correction baseline is determined according to the gestational age, for example, the fasting blood glucose value of the day is directly used in the early pregnancy, and the average value of the fasting value and the fasting value of the previous day is taken to eliminate the interference of daily fluctuation in the middle and late pregnancy, the data segment from the postprandial blood glucose peak point to the first falling of blood glucose to the correction baseline is intercepted, the least square method is used to fit a linear equation, and the absolute value of the slope is taken as the descending slope of a single meal. The final postprandial blood glucose descending slope is obtained by weighted fusion of fitting error, for example, the weight of the meal decreases by 20% for every 0.1 increase in error, and the postprandial blood glucose descending slope is obtained by inversely weighting the seven-day data according to the daily average fitting error.

[0026] Specifically, in step S2, the calculation step of the body weight gain deviation degree comprises the following steps: Obtain the body weight records of the last week from the current day in the pregnancy monitoring data set, and calculate the actual body weight growth rate in the last week from the current day; According to the gestational metabolic stage of the current gestational age, the recommended body weight growth rate in the clinical guidelines is obtained, and the body weight gain deviation degree is calculated in combination with the actual body weight growth rate.

[0027] In the implementation process, based on the fasting body weight records of the last seven days, the actual weekly body weight growth rate is calculated by linear regression, and the unit is kilogram per week. According to the recommended growth rate standard in the clinical guidelines according to the current gestational age: for example, 0.5 kg per week in early pregnancy, 0.4 kg per week in mid-pregnancy, and 0.3 kg per week in late pregnancy. The deviation degree calculation formula is actual growth rate minus recommended growth rate divided by recommended growth rate multiplied by 100%. In addition, when the daily body weight gain is more than 0.5 kg, the edema correction protocol is automatically triggered: if the user confirms that there is edema on the day, the body weight of the day is replaced by the average value of the previous three days, so as to eliminate the evaluation distortion caused by non-fat weight gain.

[0028] Step S3: based on the fasting blood glucose coefficient of variation, the postprandial blood glucose descending slope and the body weight gain deviation degree, a dynamic linkage matrix is constructed, and the risk coefficient and the linkage factor are calculated in combination with the gestational metabolic stage of the current gestational age.

[0029] Specifically, in step S3, the structure of the dynamic linkage matrix is: The first row of the matrix is the fasting blood glucose coefficient of variation, the postprandial blood glucose descending slope and the body weight gain deviation degree of the current day. The second row of the matrix is the daily change amount of the fasting blood glucose coefficient of variation, the daily change amount of the postprandial blood glucose descending slope and the daily change amount of the body weight gain deviation degree.

[0030] Specifically, in step S3, the fasting blood glucose variation coefficient, the postprandial blood glucose reduction slope and the weight gain deviation are weighted and fused according to the pregnancy metabolic stage of the current gestational week to obtain a risk coefficient.

[0031] Specifically, in step S3, the calculation step of the linkage factor includes: The overall fluctuation intensity can be obtained by calculating the Frobenius norm based on all elements of the dynamic linkage matrix; The difference between the mean postprandial blood glucose values ​​of all meals on the current day and the previous day is calculated and combined with the sign function to obtain the postprandial blood glucose change direction mark value. The overall fluctuation intensity and the postprandial blood glucose change direction mark value are weighted and fused to generate a linkage factor.

[0032] In specific implementation, a dynamic linkage matrix and a two-factor collaborative calculation mechanism were constructed to enable real-time interactive analysis and risk quantification of multidimensional metabolic characteristics. First, based on three core eigenvalues: fasting glucose coefficient of variation, postprandial glucose slope, and weight gain deviation, a two-row, three-column dynamic linkage matrix was constructed. The first row of the matrix records the daily eigenvalue, and the second row calculates the daily variation of the corresponding eigenvalue (i.e., the difference between the current value and the previous value). This matrix is ​​updated daily and retains the most recent seven days of data to ensure the complete preservation of historical fluctuations. The risk coefficient is calculated using a gestational age-dependent weighted fusion strategy. For example, in the first trimester, the fasting glucose coefficient of variation is weighted 60%, the postprandial glucose slope is weighted 30%, and the weight gain deviation is weighted 10%. In the second trimester, the weights are adjusted to 40%, 40%, and 20%. In the third trimester, the weights are assigned to weight management, with weights of 20%, 30%, and 50%. The final weighted sum is normalized to generate a risk coefficient ranging from 0 to 1, reflecting the long-term metabolic trend adapted to the gestational age. The generation of linkage factors focuses on short-term coordinated fluctuations. First, the Frobenius norm of all elements of the matrix is ​​calculated as the overall fluctuation intensity to characterize the degree of coordinated abnormality of the three eigenvalues. Then, the arithmetic mean difference of postprandial blood glucose for all valid meals, breakfast, lunch, and dinner on the current day and the previous day is extracted, and converted into a directional mark value through a sign function. For example, an increase is +1, a decrease is -1, and a flat is 0. Finally, the fluctuation intensity and the direction mark are fused according to the preset fusion weight to generate a linkage factor. This factor directly drives the subsequent dynamic adjustment of thresholds and emergency response, and transforms isolated metabolic indicators into a risk assessment dual engine with temporal correlation and pathological directionality, breaking through the fundamental defect that traditional static scoring models cannot capture the coordinated mutation of multiple indicators.

[0033] Step S4: Obtain the fasting blood glucose baseline threshold and postprandial blood glucose baseline threshold for the current gestational age from the clinical guideline library, adjust the fasting blood glucose baseline threshold by the linkage factor, and adjust the postprandial blood glucose baseline threshold by the linkage factor and weight gain deviation.

[0034] Specifically, in step S4, the adjustment step of the fasting blood glucose reference threshold value and the postprandial blood glucose reference threshold value includes: The fasting blood glucose reference adjustment amount is obtained by multiplying the gestational age, the linkage factor and the preset fasting blood glucose reference adjustment coefficient, and the adjusted fasting blood glucose reference threshold value is obtained by multiplying the fasting blood glucose reference threshold value by 1 plus the fasting blood glucose adjustment coefficient. The postprandial blood glucose reference adjustment amount is obtained by multiplying the body weight gain deviation, the linkage factor and the preset postprandial blood glucose reference adjustment coefficient, and the adjusted postprandial blood glucose reference threshold value is obtained by multiplying the postprandial blood glucose reference threshold value by 1 plus the postprandial blood glucose adjustment coefficient.

[0035] In the specific implementation process, first, the fasting blood glucose reference threshold value and the postprandial blood glucose reference threshold value corresponding to the current gestational age are extracted from the clinical guideline library. For example, when the gestational age is 28 weeks, the fasting reference value is 5.1 mmol / L, and the postprandial reference value is 7.8 mmol / L. For the fasting blood glucose threshold value, the current gestational age value, the linkage factor and the preset fasting blood glucose reference adjustment coefficient 0.02 are multiplied to generate the fasting blood glucose reference adjustment amount. The adjusted threshold value = reference threshold value x (1 + adjustment amount) is obtained by the formula to achieve multiplicative adjustment, which ensures that the threshold value is automatically tightened when the metabolic burden increases due to the increase of gestational age. For the postprandial blood glucose threshold value, the body weight gain deviation is introduced as a nutritional metabolism state correction term, which is multiplied by the linkage factor and the preset postprandial blood glucose reference adjustment coefficient 0.1 to obtain the postprandial blood glucose reference adjustment amount. Similarly, the adjusted threshold value = reference threshold value x (1 + adjustment amount) is obtained by the formula to complete the adjustment, so that the postprandial threshold value responds to short-term blood glucose fluctuations and long-term nutritional imbalance. In addition, the adjustment process has built-in safety boundary constraints. For example, the fasting adjustment threshold value is limited to the clinical safety interval of 4.0 to 5.6 mmol / L in the early gestational period and 4.3 to 5.8 mmol / L in the middle and late gestational period, and the postprandial adjustment threshold value is limited to the range of 6.0 to 10.0 mmol / L. When the boundary is exceeded, it is automatically truncated to the critical value and an alarm log is generated. Finally, the static clinical guidelines are converted into a live threshold system that adapts to individual metabolic dynamics, solving the evaluation failure problem of traditional fixed thresholds in scenarios such as increased insulin resistance in the late gestational period and abnormal weight gain.

[0036] Step S5: generating a multi-level risk assessment result based on the adjusted fasting blood glucose reference threshold value and the postprandial blood glucose reference threshold value and combining the risk coefficient.

[0037] Specifically, step S5 further includes an emergency triggering mechanism based on the linkage factor and the preset emergency triggering threshold value, including: When the linkage factor exceeds the preset emergency triggering threshold value, the risk assessment result is forcibly covered as the highest risk and the preset emergency response measure is started.

[0038] In the implementation process, first, based on the adjusted fasting blood glucose reference threshold and postprandial blood glucose reference threshold, the deviation degree of the measured fasting blood glucose value is calculated, that is, the measured value minus the adjusted threshold divided by the adjusted threshold multiplied by 100%, and the deviation degree of the measured postprandial blood glucose peak value is calculated, the calculation logic is the same as that of the fasting blood glucose deviation degree, then combined with the risk coefficient, a multi-level risk judgment is performed, for example, a three-level risk judgment mechanism is adopted, when the fasting and postprandial deviation degrees are both less than 5% and the risk coefficient is lower than 0.4, it is judged as low risk; any deviation degree reaches 5% or the risk coefficient is in the interval of 0.4 to 0.7, it is judged as medium risk; any deviation degree exceeds 15% or the risk coefficient is greater than or equal to 0.7, it is judged as high risk, and the specific exceeding index label is output, for example, "abnormal fasting blood glucose" or "deteriorating metabolic trend". On this basis, the system runs an emergency triggering mechanism in parallel: when the linkage factor exceeds the preset emergency triggering threshold 1.5, the aforementioned evaluation result is immediately forced to cover the highest risk level, and a multi-level emergency response is activated synchronously, for example, a three-level emergency response is also adopted, a red alarm containing emergency contacts and navigation of the nearest emergency agency is pushed to the pregnant woman's mobile terminal; the dynamic blood glucose monitor is automatically started in the fifteen-minute continuous sampling mode; based on the real-time data stream, the ketone body generation risk value is calculated, the calculation formula is blood glucose mean multiplied by 0.4 plus blood glucose standard deviation multiplied by 0.3 plus linkage factor multiplied by 0.3, when the ketone body generation risk value is greater than 2.0, a ketosis disposal plan containing the pregnant woman's positioning and metabolic parameters is transmitted to the associated emergency center. Through the above steps, the scientific nature of daily risk management is ensured, and the absolute priority of life safety in emergency scenarios is ensured, completely solving the inherent defects of the traditional evaluation system in risk grading ambiguity and emergency response lag.

[0039] The working principle of the dynamic risk assessment method for gestational diabetes mellitus provided by the application is as follows: This invention constructs a gestational age-adaptive multidimensional metabolic dynamic coupling model and a dual-track decision-making mechanism, achieving precise quantification and real-time response to gestational diabetes risk through spatiotemporal correlation analysis. First, it integrates multi-source monitoring data, including the current gestational age, structured fasting and postprandial blood glucose records, morning fasting weight, and estimated carbohydrate intake. Gestational age is divided into three stages according to international guidelines: early pregnancy (1-12 weeks), mid-pregnancy (13-27 weeks), and late pregnancy (28-40 weeks), and differentiated calculation rules are assigned. Based on this data, three core metabolic features are extracted in parallel: the coefficient of variation of fasting blood glucose, the slope of postprandial blood glucose decline, and the deviation of weight gain. These features are input into a dynamic linkage matrix for collaborative analysis: the first row of the matrix records the measured feature values ​​for the day, and the second row calculates the daily variation. The risk coefficient is derived by differentially weighted fusion of the three core metabolic features. The Frobenius norm of all matrix elements is used to analyze the synergistic fluctuation strength of the three features. Simultaneously, the difference between the mean blood glucose values ​​for all valid meals on the current day and the previous day is extracted and converted into directional markers using a sign function. These two features are then fused according to preset weights to generate a linkage factor. The linkage factor drives the dynamic reconstruction of the blood glucose threshold. Ultimately, the conventional path calculates the measured blood glucose deviation based on the adjusted threshold and performs multi-level risk judgment in combination with the risk coefficient. At the same time, a linkage factor hard interception mechanism is established. When the linkage factor exceeds the preset threshold of 1.5, the assessment result is forced to be the highest risk level, and multiple emergency protocols are activated simultaneously, eventually forming a full-link dynamic assessment system covering data collection, feature extraction, collaborative modeling, threshold reconstruction, hierarchical decision-making, and emergency circuit breaking.

[0040] Example 2 A dynamic risk assessment system for gestational diabetes mellitus is applied to a dynamic risk assessment method for gestational diabetes mellitus. In the specific implementation process, Figure 2 As shown, it shows a module structure diagram of a dynamic risk assessment system for gestational diabetes, including: The data acquisition module 100 is used to obtain a pregnancy monitoring data set of a pregnant woman, wherein the pregnancy monitoring data set includes the current gestational age, fasting blood glucose value record, postprandial blood glucose value record, weight record and daily carbohydrate intake estimate; The metabolic feature dynamic extraction module 200 is used to determine the pregnancy metabolic stage of the current gestational week based on the pregnancy metabolic stage divided by gestational week, and calculate the fasting blood glucose coefficient of variation, postprandial blood glucose decline slope, and weight gain deviation based on the pregnancy monitoring data set and the pregnancy metabolic stage of the current gestational week; A collaborative matrix construction module 300 is used to construct a dynamic linkage matrix based on the coefficient of variation of fasting blood glucose, the slope of postprandial blood glucose decrease, and the deviation of weight gain, and calculate the risk coefficient and linkage factor in combination with the pregnancy metabolic stage of the current gestational week; The blood glucose threshold dynamic adjustment module 400 is configured to obtain the fasting blood glucose reference threshold and the postprandial blood glucose reference threshold of the current gestational age from the clinical guideline library, and adjust the fasting blood glucose reference threshold through the linkage factor, and adjust the postprandial blood glucose reference threshold through the linkage factor and the body weight gain deviation; The multi-level risk assessment module 500 is configured to generate a multi-level risk assessment result based on the adjusted fasting blood glucose reference threshold and the postprandial blood glucose reference threshold and in combination with the risk coefficient.

[0041] The working principle of the dynamic risk assessment system for gestational diabetes mellitus provided by the application is as follows: The application acquires the gestational age, blood glucose, body weight and carbohydrate intake data in real time through the data acquisition module 100, performs multi-source verification, time sequence alignment and abnormal filtering. The metabolic characteristic dynamic extraction module 200 divides the metabolic stage according to the gestational age, and calculates the fasting blood glucose variation coefficient, the postprandial blood glucose decline slope and the body weight gain deviation in parallel. The collaborative matrix construction module 300 organizes the three characteristics into a dynamic matrix: the first row stores the daily value, and the second row records the daily change amount. The overall fluctuation intensity is calculated based on the matrix Frobenius norm, the linkage factor is generated by combining the postprandial blood glucose daily average value change direction marker and the preset weight, and the risk coefficient is output by weighting according to the gestational age difference. The blood glucose threshold dynamic adjustment module 400 calls the clinical reference threshold and implements double-path adjustment. The fasting blood glucose reference threshold is multiplied by the adjustment of “reference value x (1+gestational age x linkage factor x 0.02)”, and the postprandial blood glucose reference threshold is adjusted in cooperation by “reference value x (1+body weight deviation x linkage factor x 0.1)”, and a clinical safety boundary constraint is applied. The multi-level risk assessment module 500 implements double-track decision: the conventional path calculates the blood glucose deviation based on the adjusted threshold, and performs multi-level risk judgment in combination with the risk coefficient, and an emergency triggering mechanism based on the linkage factor and the preset emergency triggering threshold is also added. When the linkage factor exceeds the preset emergency triggering threshold, the high-risk result is forcibly overwritten, and the alarm push, minute-level monitoring and emergency linkage are activated. Through the closed loop link of data governance, characteristic extraction, matrix modeling, threshold reconstruction and hierarchical decision, the application realizes the intelligentization of dynamic risk dynamic perception and response of gestational diabetes mellitus.

[0042] The application is described with reference to flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams can be implemented by computer program instructions, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a machine that implements the flowcharts and / or block diagrams. Figure 1 The flow or flows and / or blocksFigure 1 means for performing the function specified by the block or blocks.

[0043] Those skilled in the art can understand that all or part of the steps in the above-mentioned embodiments can be completed by instructing the relevant hardware through a program, and the program can be stored in a computer readable storage medium, including Read-Only Memory (ROM), Random Access Memory (RAM), Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only Memory (EPROM), One-time Programmable Read-Only Memory (OTPROM), Electrically-Erasable Programmable Read-Only Memory (EEPROM), Compact Disc Read-Only Memory (CD-ROM) or other optical disk storage, magnetic disk storage, magnetic tape storage, or any other medium that can be used to carry or store data which can be read by a computer.

[0044] It should also be noted that the terms "comprising", "containing", or any other variant thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements, but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the statement "comprising a" does not exclude the existence of additional identical elements in the process, method, article, or apparatus that includes the element.

Claims

1. A method for dynamic risk assessment of gestational diabetes mellitus, characterized in that: The dynamic risk assessment method comprises the following steps: S1: Obtaining a pregnancy monitoring dataset of a pregnant woman, wherein the pregnancy monitoring dataset includes the current gestational age, fasting blood glucose value records, postprandial blood glucose value records, weight records, and daily carbohydrate intake estimates; S2: Determine the current gestational metabolic stage based on the gestational metabolic stages divided by gestational age. Calculate the fasting blood glucose coefficient of variation, postprandial blood glucose decline slope, and weight gain deviation based on the pregnancy monitoring data set and the current gestational metabolic stage. S3: A dynamic linkage matrix is ​​constructed based on the coefficient of variation of fasting blood glucose, the slope of postprandial blood glucose decrease, and the deviation of weight gain, and the risk coefficient and linkage factor are calculated in combination with the pregnancy metabolic stage of the current gestational age; S4: Obtain the fasting blood glucose baseline threshold and postprandial blood glucose baseline threshold for the current gestational age from the clinical guideline database, adjust the fasting blood glucose baseline threshold by the linkage factor, and adjust the postprandial blood glucose baseline threshold by the linkage factor and weight gain deviation; S5: Generate a multi-level risk assessment result based on the adjusted fasting blood glucose baseline threshold and postprandial blood glucose baseline threshold and combined with the risk factor.

2. A method for dynamic risk assessment of gestational diabetes mellitus according to claim 1, characterized in that: In step S2, the calculation of the coefficient of variation of fasting blood glucose includes the following steps: Obtain fasting blood glucose records from the pregnancy monitoring data set within the week before the current day and calculate the fasting blood glucose standard deviation and fasting blood glucose mean; Based on the current gestational age and the gestational metabolic stage, the preset coefficient of variation formula is selected to process the fasting blood glucose standard deviation and the fasting blood glucose mean to obtain the fasting blood glucose coefficient of variation.

3. A method for dynamic risk assessment of gestational diabetes mellitus according to claim 2, characterized in that: In step S2, the calculation of the postprandial blood glucose reduction slope includes the following steps: The postprandial blood glucose value records for the week preceding the current day were obtained from the pregnancy monitoring data set, and multiple postprandial blood glucose curves for each day were obtained through interpolation. The postprandial blood glucose peak of each postprandial blood glucose curve was identified, and the fasting blood glucose corrected baseline was determined based on the current gestational metabolic stage and daily fasting blood glucose records; The data segment of each postprandial blood glucose curve from the postprandial blood glucose peak point to the first time the blood glucose value is less than or equal to the corrected baseline was intercepted. The linear regression equation was fitted using the least squares method and the absolute value of the regression slope was taken as the blood glucose drop slope of the single meal. The fitting error was also calculated. Based on the fitting error, the fusion weight of the blood glucose reduction slope of a single meal is determined and weighted fusion is performed to obtain the daily postprandial blood glucose reduction slope. The daily postprandial blood glucose reduction slope is weightedly fused to obtain the postprandial blood glucose reduction slope.

4. A method for dynamic risk assessment of gestational diabetes mellitus according to claim 3, characterized in that: In step S2, the step of calculating the weight gain deviation includes the following steps: Obtain the weight records for the week preceding the current day from the pregnancy monitoring data set, and calculate the actual weight gain within the week preceding the current day; The recommended weight growth rate in clinical guidelines is obtained based on the current gestational metabolic stage, and the weight gain deviation is calculated based on the actual weight growth rate.

5. A method for dynamic risk assessment of gestational diabetes mellitus according to claim 4, characterized in that: In step S3, the structure of the dynamic linkage matrix is: The first row of the matrix is ​​the coefficient of variation of fasting blood glucose, slope of postprandial blood glucose decrease and deviation of weight gain on that day; The second row of the matrix is ​​the daily change in the coefficient of variation of fasting blood glucose, the daily change in the slope of postprandial blood glucose decrease, and the daily change in the deviation of weight gain.

6. A method for dynamic risk assessment of gestational diabetes mellitus according to claim 5, characterized in that: In step S3, the fasting blood glucose variation coefficient, the postprandial blood glucose reduction slope and the weight gain deviation are weighted and fused according to the pregnancy metabolic stage of the current gestational week to obtain a risk coefficient.

7. A method for dynamic risk assessment of gestational diabetes mellitus according to claim 6, characterized in that: In step S3, the calculation steps of the linkage factor include: The overall fluctuation intensity can be obtained by calculating the Frobenius norm based on all elements of the dynamic linkage matrix; The difference between the mean postprandial blood glucose values ​​of all meals on the current day and the previous day is calculated and combined with the sign function to obtain the postprandial blood glucose change direction mark value. The overall fluctuation intensity and the postprandial blood glucose change direction mark value are weighted and fused to generate a linkage factor.

8. A method for dynamic risk assessment of gestational diabetes mellitus according to claim 7, characterized in that: In step S4, the steps of adjusting the fasting blood glucose reference threshold and the postprandial blood glucose reference threshold include: The fasting blood glucose baseline adjustment amount is obtained by multiplying the gestational age, linkage factor and the preset fasting blood glucose baseline adjustment coefficient, and the fasting blood glucose baseline threshold is multiplied by 1 and added to the fasting blood glucose adjustment coefficient to obtain the adjusted fasting blood glucose baseline threshold; The postprandial blood glucose baseline adjustment amount is obtained by multiplying the weight gain deviation, the linkage factor and the preset postprandial blood glucose baseline adjustment coefficient. The adjusted postprandial blood glucose baseline threshold is obtained by multiplying the postprandial blood glucose baseline threshold by 1 and adding the postprandial blood glucose adjustment coefficient.

9. A method for dynamic risk assessment of gestational diabetes mellitus according to claim 8, characterized in that: Step S5 also includes an emergency trigger mechanism based on the linkage factor and a preset emergency trigger threshold, including: When the linkage factor exceeds the preset emergency trigger threshold, the risk assessment result is forced to be the highest risk and the preset emergency response measures are initiated.

10. A dynamic risk assessment system for gestational diabetes mellitus, characterized in that: A method for dynamic risk assessment of gestational diabetes mellitus according to any one of claims 1 to 9, wherein the dynamic risk assessment system comprises: A data acquisition module is used to obtain a pregnancy monitoring data set of a pregnant woman, wherein the pregnancy monitoring data set includes the current gestational age, fasting blood glucose value record, postprandial blood glucose value record, weight record and daily carbohydrate intake estimate; The metabolic feature dynamic extraction module is used to determine the current gestational metabolic stage based on the gestational metabolic stages divided by gestational age. It also calculates the fasting blood glucose coefficient of variation, postprandial blood glucose decline slope, and weight gain deviation based on the pregnancy monitoring dataset and the current gestational metabolic stage. A collaborative matrix construction module is used to construct a dynamic linkage matrix based on the coefficient of variation of fasting blood glucose, the slope of postprandial blood glucose decline, and the deviation of weight gain, and calculate the risk coefficient and linkage factor in combination with the pregnancy metabolic stage of the current gestational week; The blood glucose threshold dynamic adjustment module is used to obtain the fasting blood glucose baseline threshold and postprandial blood glucose baseline threshold for the current gestational week from the clinical guideline library, and adjust the fasting blood glucose baseline threshold by the linkage factor, and adjust the postprandial blood glucose baseline threshold by the linkage factor and weight gain deviation; The multi-level risk assessment module is used to generate a multi-level risk assessment result based on the adjusted fasting blood glucose baseline threshold and postprandial blood glucose baseline threshold and in combination with the risk factor.

Images