Epidural delivery analgesic lower head dystocia risk prediction method and system

Abstract

The invention discloses an epidural delivery analgesic lower head dystocia risk prediction method and system. The method comprises the following steps: retrospectively collecting clinical feature data of a parturient to be parturified under epidural delivery analgesia; the method comprises the following steps: determining classification tangency points of continuous variables according to clinical medicine consensus, literature review or statistical distribution, converting the continuous variables into ordered or disordered classification variables, performing code conversion on the collected classification variables, and constructing a structured data set; carrying out preliminary screening on all feature data by using single-factor logistic regression analysis and carrying out multi-collinearity diagnosis; the screened features are incorporated into a multi-factor logistic regression model for optimization to determine key prediction variables, and a logistic regression prediction model is constructed to derive a logistic regression equation; and converting the logistic regression model into a Nomogram column graph model, and outputting a prediction result of the occurrence risk probability of the head dystocia. According to the scheme, early risk prediction of epidural delivery analgesia lower head dystocia is realized, and an earlier decision window is provided for clinic.

Description

Technical Field

[0001] This invention relates to the field of medical risk prediction technology, and in particular to a method and system for predicting the risk of cephalic presentation dystocia under epidural analgesia. Background Technology

[0002] Cephalic presentation dystocia, the most common type of dystocia, refers to a mismatch between the fetal head position and the maternal pelvis. This is often manifested as obstructed fetal head rotation, poor flexion, and arrested descent, requiring assisted delivery or cesarean section. This type of dystocia is usually diagnosed in the late active phase or the second stage of labor and significantly increases the risk of maternal and infant complications, including postpartum hemorrhage, infection, fetal distress, negative labor experiences, and even maternal death. According to the World Health Organization (WHO), the global cesarean section rate has reached 21% and is continuously rising. Dystocia is one of the important indications for cesarean section during labor, and it is associated with approximately 8% of maternal deaths. Statistics show that about 10% of term pregnancies with cephalic presentation experience complications during delivery, of which approximately 20%–25% require cesarean section.

[0003] Epidural analgesia for labor is widely recommended by the World Health Organization (WHO) as the preferred method for relieving labor pain due to its high efficacy and safety. It is widely used globally. However, the widespread adoption of epidural analgesia for labor presents new challenges to traditional labor management models. There is no complete consensus on the impact of epidural analgesia on the labor process; previous studies have shown that it may prolong the first and second stages of labor, increase the risk of cesarean delivery, maternal fever, and neonatal intensive care unit admission.

[0004] While epidural analgesia relieves pain, it may weaken the early clinical signs of dystocia, thus affecting the timely identification and intervention of dystocia. This not only increases the risk of prolonged labor, emergency cesarean section, and difficult instrumental delivery, but may also lead to maternal soft tissue injury, bleeding, infection, and adverse outcomes such as fetal distress, neonatal asphyxia, and acidosis. Existing predictive models have significant limitations: First, their construction largely relies on data from non-analgesic populations, resulting in limited applicability under epidural analgesia; second, they generally focus on predicting conversion to cesarean section, lacking an early identification system for the specific process of cephalic presentation dystocia; furthermore, most models fail to effectively integrate labor progress indicators, making it difficult to reflect the actual progress of the delivery process. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a method and system for predicting the risk of cephalic presentation dystocia under epidural analgesia, specifically designed for early prediction of cephalic presentation dystocia under epidural analgesia and capable of integrating labor progress indicators.

[0006] This invention is achieved using the following technical solution: Firstly, a method for predicting the risk of cephalic presentation dystocia under epidural analgesia includes the following steps: Step S1: Retrospectively collect clinical characteristic data of women in labor who were under epidural analgesia and classify them into continuous variables and categorical variables; Step S2: Feature data preprocessing. Based on clinical medical consensus, literature review or statistical distribution, determine the classification cut-off points of continuous variables, convert continuous variables into ordered or unordered categorical variables, encode and transform the collected categorical variables, and construct a structured dataset. Step S3: Use univariate logistic regression analysis to initially screen all preprocessed feature data, retain features related to cephalic presentation dystocia, and perform multicollinearity diagnosis on the screened feature data; Step S4: Incorporate the selected features into the multi-factor logistic regression model, use the stepwise regression method to further optimize the feature combination, finally determine the key predictive variables, and construct a logistic regression prediction model based on the regression coefficients of the variables, and derive the corresponding logistic regression equation. Step S5: Transform the logistic regression model into a Nomogram model for individualized risk prediction at the beginning of the active clinical phase. Output the predicted probability of cephalic presentation dystocia by using the total score of the nomogram to correspond to the risk probability.

[0007] Specifically, the continuous variables collected in step S1 include maternal age, BMI (maternal prenatal body mass index), fetal abdominal circumference, and gestational age; the categorical variables collected include parity, cervical edema, induction of labor, prolonged latent period, nuchal cord, premature rupture of membranes, fetal presentation at the beginning of the active phase, and fetal position.

[0008] Specifically, both the univariate logistic regression analysis and the multivariate logistic regression analysis were implemented using R language and SPSS software, and the stepwise regression method was based on the AIC criterion for variable selection.

[0009] Specifically, the feature related to cephalic presentation dystocia in step S3 is manifested as follows: the probability P < 0.1 in univariate logistic regression analysis.

[0010] Specifically, the key predictive variables finally determined in step S4 are seven in total, including: maternal age, parity, premature rupture of membranes, fetal abdominal circumference, prolonged latent phase, fetal presenting part and fetal position at the beginning of the active phase.

[0011] Specifically, the logistic regression equation derived in step S4 is expressed as follows: ; in, This indicates the probability of cephalic presentation dystocia. For the intercept term, For the first The regression coefficients of the predictor variables, For the values ​​that the corresponding variables can take, .

[0012] Specifically, the transformation of the Nomogram model in step S5 is as follows: based on the regression coefficients of each variable in the logistic regression equation derived from the multi-factor logistic regression model, the values ​​of each variable are mapped to scores, and the total score corresponds to the predicted probability.

[0013] Specifically, step S5 further includes a Nomogram model verification step: internal verification is performed using the Bootstrap repeated sampling test method to evaluate its discrimination and calibration.

[0014] On the other hand, a risk prediction system for cephalic presentation dystocia under epidural analgesia predicts the risk of cephalic presentation dystocia in expectant mothers under epidural analgesia and outputs visualized results, including the following modules: Data acquisition module: Connects to the hospital's medical information system to obtain clinical data of expectant mothers; Data processing module: performs data preprocessing and variable transformation; Feature selection module: Performs univariate and multivariate logistic regression to analyze the data; Model building module: Builds logistic regression models and nomogram models; Risk output module: Through a nomogram model, it outputs individualized risk values ​​for cephalic presentation dystocia at the beginning of the active phase, allowing medical staff to intervene and manage individualized delivery based on the risk prediction results.

[0015] The beneficial effects of this invention are as follows: This method predicts the risk of cephalic presentation dystocia in women undergoing epidural analgesia by constructing a nomogram model containing seven objective clinical variables. The model has good discriminative ability, with parity, prolonged latent phase, and fetal position at the onset of the active phase being key predictive factors. This method compensates for the weakening of abnormal labor signals that may be caused by epidural analgesia, providing clinicians with an early and objective risk assessment method. This facilitates more timely intervention and individualized labor management, potentially improving maternal and infant outcomes. Early risk prediction provides clinicians with an earlier decision-making window to implement targeted interventions such as position adjustment and manual rotation of the fetal head, reducing the need for cesarean section and improving maternal and infant safety. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0017] Figure 1 This is a basic flowchart for predicting the risk of cephalic presentation dystocia under epidural analgesia in an embodiment of the present invention; Figure 2 This is a schematic diagram of the Nomogram model in an embodiment of the present invention; Figure 3 This is a schematic diagram of the ROC curve of the Nomogram model in an embodiment of the present invention; Figure 4 This is a schematic diagram of the calibration of the Nomogram model in an embodiment of the present invention; Figure 5 This is a schematic diagram of a clinical decision-making model based on a Nomogram nomogram in an embodiment of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0019] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0020] The following is in conjunction with the appendix Figures 1-5 The following describes some embodiments of the present invention in detail. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0021] This invention proposes a method for predicting the risk of cephalic presentation dystocia under epidural analgesia. In a preferred embodiment, the specific steps of this method are as follows: Figure 1 ,include: Step 1: Data collection. Retrospectively collect clinical characteristic data of the mother, including maternal age, maternal prenatal body mass index (BMI), fetal abdominal circumference, parity, nuchal cord, cervical edema, prolonged latent period, induction of labor, gestational age, premature rupture of membranes, fetal presentation and fetal position at the beginning of the active phase. Step 2: Data preprocessing. Collected continuous variables (including maternal age, BMI, fetal abdominal circumference, and gestational age) are preprocessed by determining classification cutoff points based on clinical consensus, literature review, or statistical distribution, and then converted into ordered or unordered categorical variables. For inherent categorical variables (such as parity, cervical edema, induction of labor, prolonged latent period, nuchal cord, premature rupture of membranes, fetal presentation at the beginning of the active phase, and fetal position), encoding and transformation are performed to construct a structured dataset. Step 3: Feature screening. Univariate logistic regression analysis was used to initially screen all features, retaining those related to cephalic presentation dystocia (P < 0.1). Multicollinearity was diagnosed on the screened features to confirm that there were no significant collinearity issues. Step 4: Model Construction. The selected features are incorporated into a multi-factor logistic regression model. Stepwise regression is used to further optimize the feature combination, ultimately identifying seven key predictor variables. Based on the regression coefficients of these seven variables, a logistic regression prediction model is constructed, and the corresponding logistic regression equation is derived. ;in, This indicates the probability of cephalic presentation dystocia. For the intercept term, For the first The regression coefficients of the predictor variables, For the values ​​that the corresponding variables can take, ; Step 5: Model visualization and prediction. The logistic regression model is transformed into a nomogram model for individualized risk prediction at the beginning of the active clinical phase (cervical dilation 4-6cm). The risk probability corresponding to the total score of the nomogram is used to output the probability of cephalic presentation dystocia and the prediction result of the risk of cephalic presentation dystocia.

[0022] In this embodiment, univariate logistic regression and multivariate logistic regression analyses are implemented using R language and SPSS software. The stepwise regression method selects variables based on the AIC criterion. The nomogram model is constructed based on the regression coefficients of each variable in the multivariate logistic regression, mapping the values ​​of each variable to scores, with the total score corresponding to the predicted probability, thus achieving risk visualization.

[0023] In this embodiment, the seven key predictive variables include: maternal age, parity, premature rupture of membranes, fetal abdominal circumference, prolonged latent phase, presenting part at the beginning of the active phase, and fetal position. The logistic regression equation based on 7 key predictor variables is as follows: logit(P) = -8.079 + 1.004 × premature rupture of membranes (0 for No, 1 for Yes) + 2.626 × prolonged latency period (0 for No, 1 for Yes) + 0.346 × fetal abdominal circumference (0 for <330mm, 1 for 330-350) + 1.634 × fetal abdominal circumference (0 for <330mm, 1 for >350) + 0.762 × maternal age (0 for <25, 1 for 25-30) + 1.352 × maternal age (0 for <25, 1 for >30) + 1.308 × fetal presentation at the beginning of active phase (0 for >-1, 1 for ≤-1) + 2.128 × fetal position at the beginning of active phase (0 for OA, 1 for NotOA) + 2.782 × parity (0 for ≠1, 1 for =1).

[0024] In this embodiment, the nomogram model also includes a model validation step: internal validation is performed using the Bootstrap repeated sampling test method to evaluate its discrimination (e.g., AUC value) and calibration (e.g., Hosmer-Lemeshow test).

[0025] This invention also proposes a risk prediction system for cephalic presentation dystocia under epidural analgesia, which can be configured in a hospital management system to predict the risk of cephalic presentation dystocia in expectant mothers under epidural analgesia and output visualized results, including the following modules: Data acquisition module: Connects to the hospital's medical information system to obtain clinical data of expectant mothers; Data processing module: performs data preprocessing and variable transformation; Feature selection module: Performs univariate and multivariate logistic regression to analyze the data; Model building module: Builds logistic regression models and nomogram models; Risk output module: Through a nomogram model, it outputs individualized risk values ​​for cephalic presentation dystocia at the beginning of the active phase, allowing medical staff to intervene and manage individualized delivery based on the risk prediction results.

[0026] This system is based on a method for predicting the risk of cephalic presentation dystocia under epidural analgesia. It acquires and analyzes data from pregnant women who are undergoing epidural analgesia and are in labor. After calculation by the system, the prediction results are output. Medical staff can then make decisions based on these prediction results and implement targeted interventions such as position adjustment and manual rotation of the fetal head, ultimately improving delivery outcomes, reducing cesarean sections, and enhancing maternal and infant safety.

[0027] The following is a detailed explanation of this protocol using specific data. This embodiment is a retrospective cohort study based on medical records from the hospital's obstetrics department from January 2021 to October 2024. The study subjects were full-term, singleton, cephalic presentation pregnant women who underwent epidural analgesia and vaginal delivery. Inclusion criteria were: full-term (37-41+6 weeks), single live fetus, cephalic presentation, use of epidural analgesia, and no contraindications to vaginal delivery. Exclusion criteria were: pregnant women with mental disorders or cognitive impairment, cervical dilation >6cm at admission, those who underwent cesarean section due to poor fetal condition or fetal distress, those who requested cesarean section during the trial of labor due to social factors, and those with incomplete general information.

[0028] In this embodiment, the specific content of the risk prediction for cephalic presentation dystocia under epidural analgesia includes, in sequence: I. Data Collection Based on a literature review and expert discussions on high-risk factors for cephalic presentation dystocia, the following medical records were retrospectively collected: baseline data (age, height, prepartum weight, parity, gestational age) and pre-labor conditions (premature rupture of membranes, induction of labor, nuchal cord, fetal abdominal circumference) and intrapartum factors (prolonged latent phase, cervical edema, fetal presentation at the onset of the active phase, and fetal position). A binary variable, "whether cephalic presentation dystocia occurred," was used as the outcome variable (assigned values: 1 = yes, 0 = no).

[0029] In this embodiment, the relevant variables are defined as follows: Fetal abdominal circumference: Measured directly via ultrasound within one week before delivery by taking a transverse section of the abdomen perpendicular to the fetal spine at the point where the umbilical vein enters the liver. Cervical edema: Its assessment is performed by an experienced midwife or obstetrician who palpates the cervix at the beginning of the active phase to assess its texture, thickness, and elasticity. In some cases, ultrasound is used to measure the thickness or echogenicity of the cervical tissue to increase objectivity.

[0030] II. Statistical Analysis Statistical analysis was performed using R 4.4.3 and SPSS 25.0 software. Normally distributed continuous data were described using mean ± standard deviation, non-normally distributed continuous data were described using median (interquartile range), and categorical data were described using percentages. Descriptive analysis was performed using R (tableone package).

[0031] In this embodiment, 924 cases were initially included in the screening, and 69 cases were excluded according to the inclusion and exclusion criteria. Among these, 37 cases were excluded due to missing key variables (fetal abdominal circumference in 11 cases, maternal height in 16 cases, and prenatal weight in 10 cases). A total of 855 cases were ultimately included in the analysis, with complete and unmissing data. Univariate analysis used R software to construct a logistic regression model to examine the relationship between the following variables and cephalic presentation dystocia under epidural analgesia: parity, maternal age, prenatal BMI, premature rupture of membranes, induction of labor, cervical edema, prolonged latency period, fetal abdominal circumference, nuchal cord, gestational age, active early fetal presentation, and fetal position. Results are expressed as odds ratios (OR) and 95% confidence intervals (CI). Ten variables with P < 0.1 in the univariate analysis were included in the multicollinearity diagnosis. SPSS 25.0 was used to calculate the variance inflation factor (VIF) and tolerance (TOL), and variables with VIF > 10 or TOL < 0.1 were excluded to ensure model stability. Multivariate analysis was performed using R software with backward stepwise regression. A multivariate logistic regression model was constructed from 10 variables after collinearity screening to identify independent predictors. Results were reported as odds ratios (95% CI), with P < 0.05 considered statistically significant. Predictive nomograms were constructed using the `rms` package in R software; receiver operating characteristic (ROC) curves were plotted using the `pROC` package, and the area under the curve (AUC) was calculated to assess discrimination; calibration was evaluated using the `calibrate` package and the Hosmer-Lemeshow test (P > 0.05 indicated good calibration); and decision curve analysis (DCA) was performed using the `rmda` package to assess clinical applicability. Internal validation was performed using the Bootstrap method (1000 resamplings), and the Brier score was used to evaluate prediction error (0–0.25 was acceptable, closer to 0 was better). Statistical power: The key predictors were further analyzed using the `power.prop.test` function in R. For example, regarding the association between prolonged latency and dystocia, given the observed event rate and actual sample size, the statistical power of this study was 100% at a two-tailed significance level of α=0.05. At an 80% statistical power level, this study detected an effect with an OR > 2.61.

[0032] Epidural labor analgesia protocol: A catheter is placed in the L2–3 / L3–4 interspace (depth 3–5 cm). The trial dose is 3–5 mL of 1.5% lidocaine. Continuous analgesia is provided using a 0.1% ropivacaine + fentanyl (2 μg / mL) dilution solution, with a background infusion of 6–8 mL / h and a patient-controlled booster dose of 6–8 mL (locked out for 15–20 minutes). There is no fixed cervical dilation criterion for initiating analgesia; it is determined based on the mother's pain level and the progress of labor.

[0033] In this embodiment, data from 924 pregnant women were collected from January 2021 to October 2024. All were full-term, single live fetuses in cephalic presentation who attempted vaginal delivery under epidural analgesia. Ultimately, data from 855 women were used for further modeling analysis, of which 109 cases (12.7%) were cephalic presentation dystocia and 746 cases (87.3%) were non-cephalic presentation dystocia, as shown in Table 1.

[0034] Table 1. Characteristics of the dystocia group and the non-dystocia group In this embodiment, the results of univariate analysis showed that maternal age, premature rupture of membranes, induction of labor, cervical edema, prolonged latent period, fetal abdominal circumference, parity, days of labor, presenting part and fetal position at the beginning of the active phase (P<0.1) are as shown in Table 2.

[0035] Table 2 Univariate Analysis Table In this embodiment, the multivariate regression analysis showed that seven variables were included in the model: premature rupture of membranes (OR 2.73, 95% CI 1.46–5.11, P = 0.002), prolonged latency period (OR 13.82, 95% CI 7.51–25.42, P < 0.001), fetal abdominal circumference (OR 5.13, 95% CI 2.22–11.81, P < 0.001), and maternal age (OR 3.87). The odds ratios (OR 2.82, 95% CI 1.29–11.59, P < 0.016), fetal presentation position at the beginning of the active period (OR 2.82, 95% CI 1.63–4.88, P < 0.001), fetal orientation at the beginning of the active period (OR 8.4, 95% CI 4.83–14.59, P < 0.001), and parity (OR 16.16, 95% CI 2.95–88.55, P = 0.001) are shown in Table 3.

[0036] Table 3 Multifactor Analysis Table Based on the bias regression coefficients of the independent variables in the above multivariate logistic regression equation, a risk prediction model for cephalic presentation dystocia in women undergoing epidural analgesia is constructed. The calculation is as follows: logit(P) = -8.079 + 1.004 × premature rupture of membranes (0 for No, 1 for Yes) + 2.626 × prolonged latent period (0 for No, 1 for Yes) + 0.346 × fetal abdominal circumference (0 for <330mm, 1 for 330-350) + 1.634 × fetal abdominal circumference (0 for <330mm, 1 for >350) + 0.762 × maternal age (0 for <25, 1 for 25-30) + 1.352 × maternal age (0 for <25, 1 for >30) + 1.308 × fetal presentation at the beginning of active period (0 for >-1, 1 for ≤-1) + 2.128 × fetal position at the beginning of active period (0 for OA, 1 for NotOA) + 2.782 × parity (0 for ≠1, 1 for =1).

[0037] The predictive model, presented in nomogram form, is used to predict the probability of cephalic presentation dystocia in women undergoing epidural analgesia, such as... Figure 2 As shown.

[0038] The final model's performance evaluation showed an area under the receiver operating characteristic (AUC) of 0.910 (95% CI: 0.878–0.941), and a corrected AUC of 0.911. Figure 3 The internally validated calibration curves and Hos-Mer-Lemeshow test are as follows: Figure 4 The model calibration is shown to be good (P=0.992>0.05, Brier's test=0.063<0.25). Its practicality was evaluated using DCA as follows: Figure 5 The model demonstrates practical value in the range of 5% to 83%.

[0039] In summary, this study successfully constructed and validated a nomogram model for predicting cephalic presentation dystocia in women receiving epidural analgesia. The model incorporated seven objective clinical variables from the pre-labor and intrapartum periods. Parity, prolonged latency period, and fetal position at the onset of the active phase were the most effective independent predictive factors. This model aims to provide an early and objective warning tool for the clinical manifestations of dystocia that may become elusive due to weakened pain signals under epidural analgesia.

[0040] This study first observed that parity (primiparous women) was the strongest predictor of cephalic presentation dystocia. Secondly, prolonged latency also showed strong predictive value in this model. This risk prediction model incorporates these two factors, using objective indicators to enhance the monitoring and risk identification of labor progress, especially when analgesia may weaken early abnormal signals.

[0041] Furthermore, fetal abdominal circumference, as a reliable ultrasound indicator for assessing fetal size and predicting macrosomia, is crucial for the choice of delivery method. This protocol also found that a fetal abdominal circumference >350 mm is an independent risk factor for cephalic presentation dystocia under epidural analgesia. After epidural analgesia, the severe pain caused by dystocia due to excessive fetal abdominal circumference may be reduced, potentially affecting timely identification by labor managers and delaying intervention. This protocol provides objective indicators to replace subjective symptoms as a predictive tool, effectively compensating for the potential reduction in clinical manifestations caused by epidural analgesia.

[0042] This model also integrates other important risk factors: increasing maternal age is associated with physiological decline in uterine myometrial function and a reduction in oxytocin receptors, which may affect uterine contraction efficiency and increase the risk of abnormal labor. Premature rupture of membranes not only increases the risk of amniotic cavity infection and puerperal infection, but may also affect uterine contraction rhythm and fetal rotation, thereby increasing the chance of dystocia. Incorporating these factors into the model helps to form a comprehensive risk assessment of the maternal baseline status, providing a basis for individualized labor management. This study sets the application stage of the predictive model at cervical dilation of 4-6 cm, aiming to identify risks and correct variable risk factors in a timely manner at the beginning or near the active phase, thereby providing decision-making for reversing potential adverse delivery outcomes and reflecting a prevention-oriented clinical management approach.

[0043] For the foregoing embodiments, in order to simplify the description, they are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, because according to this application, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions involved are not necessarily essential to this application.

[0044] The above embodiments describe the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Modifications and variations made by those skilled in the art without departing from the spirit and scope of the invention should be within the protection scope of the appended claims.

Claims

1. A method for predicting the risk of cephalic presentation dystocia under epidural analgesia, characterized in that, Includes the following steps: Step S1: Retrospectively collect clinical characteristic data of women in labor who were under epidural analgesia and classify them into continuous variables and categorical variables; Step S2: Feature data preprocessing. Based on clinical medical consensus, literature review or statistical distribution, determine the classification cut-off points of continuous variables, convert continuous variables into ordered or unordered categorical variables, encode and transform the collected categorical variables, and construct a structured dataset. Step S3: Use univariate logistic regression analysis to initially screen all preprocessed feature data, retain features related to cephalic presentation dystocia, and perform multicollinearity diagnosis on the screened feature data; Step S4: Incorporate the selected features into the multi-factor logistic regression model, use the stepwise regression method to further optimize the feature combination, finally determine the key predictive variables, and construct a logistic regression prediction model based on the regression coefficients of the variables, and derive the corresponding logistic regression equation. Step S5: Transform the logistic regression model into a Nomogram model for individualized risk prediction at the beginning of the active clinical phase. Output the predicted probability of cephalic presentation dystocia by using the total score of the nomogram to correspond to the risk probability.

2. The method for predicting the risk of cephalic presentation dystocia under epidural analgesia as described in claim 1, characterized in that, The continuous variables collected in step S1 include maternal age, BMI (maternal prenatal body mass index), fetal abdominal circumference, and gestational age; the categorical variables collected include parity, cervical edema, induction of labor, prolonged latent period, nuchal cord, premature rupture of membranes, fetal presentation at the beginning of the active phase, and fetal position.

3. The method for predicting the risk of cephalic presentation dystocia under epidural analgesia as described in claim 1, characterized in that, Both the univariate logistic regression analysis and the multivariate logistic regression analysis were implemented using R language and SPSS software. The stepwise regression method was based on the AIC criterion for variable selection.

4. The method for predicting the risk of cephalic presentation dystocia under epidural analgesia as described in claim 3, characterized in that, The specific characteristics related to cephalic presentation dystocia in step S3 are reflected in the following: the probability P < 0.1 in univariate logistic regression analysis.

5. The method for predicting the risk of cephalic presentation dystocia under epidural analgesia as described in claim 1, characterized in that, The key predictive variables finally determined in step S4 are specifically seven, including: maternal age, parity, premature rupture of membranes, fetal abdominal circumference, prolonged latent phase, fetal presenting part and fetal position at the beginning of the active phase.

6. The method for predicting the risk of cephalic presentation dystocia under epidural analgesia as described in claim 5, characterized in that, The logistic regression equation derived in step S4 is expressed as follows: ； in, This indicates the probability of cephalic presentation dystocia. For the intercept term, For the first The regression coefficients of the predictor variables, For the values ​​that the corresponding variables can take, .

7. The method for predicting the risk of cephalic presentation dystocia under epidural analgesia as described in claim 6, characterized in that, The transformation of the Nomogram model in step S5 is specifically as follows: based on the regression coefficients of each variable in the logistic regression equation derived from the multi-factor logistic regression model, the values ​​of each variable are mapped to scores, and the total score corresponds to the predicted probability.

8. The method for predicting the risk of cephalic presentation dystocia under epidural analgesia as described in claim 1, characterized in that, Step S5 also includes a Nomogram model verification step: internal verification is performed using the Bootstrap repeated sampling test method to evaluate its discrimination and calibration.

9. A risk prediction system for cephalic presentation dystocia under epidural analgesia, based on the risk prediction method for cephalic presentation dystocia under epidural analgesia according to any one of claims 1 to 8, predicts the risk of cephalic presentation dystocia in expectant mothers under epidural analgesia and outputs visualized results, characterized in that, Includes the following modules: Data acquisition module: Connects to the hospital's medical information system to obtain clinical data of expectant mothers; Data processing module: performs data preprocessing and variable transformation; Feature selection module: Performs univariate and multivariate logistic regression to analyze the data; Model building module: Builds logistic regression models and nomogram models; Risk output module: Through a nomogram model, it outputs individualized risk values ​​for cephalic presentation dystocia at the beginning of the active phase, allowing medical staff to intervene and manage individualized delivery based on the risk prediction results.

Images