Newborn asphyxia risk prediction model construction method and device

Abstract

The invention discloses a neonatal asphyxia risk prediction model construction method and device, which are applied to the field of computer models for disease prediction, and are used for acquiring clinical feature data of a neonatal asphyxia group and a healthy neonatal group, including pregnant mother information, fetus information and other information; performing single-factor analysis on clinical risk factors between the two groups of data, and screening out single-factor predictive variables; and by taking the screened variables as independent variables and taking whether suffocation occurs as dependent variables, carrying out binary Logistic regression analysis by adopting a forward stepwise method, and establishing a risk prediction model. A visual column diagram is drawn according to the prediction model, and the prediction model has good distinction degree and calibration degree through evaluation. According to the method, key independent risk factors are screened from multiple factors through a scientific statistical analysis method, a prediction model beneficial to early clinical recognition of neonatal suffocation high-risk groups is constructed, a basis is provided for timely making individual intervention measures, and the method is of great significance in preventing neonatal suffocation and improving the survival rate of neonates.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of computer models for disease prediction, in particular to a method and device for constructing a neonatal asphyxia risk prediction model. BACKGROUND

[0002] Neonatal asphyxia is one of the main causes of neonatal death. Neonatal asphyxia refers to a disease in which the fetus is hypoxic due to various reasons before, during or after delivery, resulting in no spontaneous breathing or failure to establish regular breathing within one minute after birth, with hypoxemia and hypercapnia as the main pathophysiological changes. Although the progress of medical level has significantly improved the survival rate of neonatal asphyxia children, many neonates still suffer from multiple organ damage due to asphyxia, causing heavy burden to families and society.

[0003] The occurrence of neonatal asphyxia is related to a series of complex factors, and it is crucial to comprehensively identify the risk factors that play an important role and intervene in advance. At present, a variety of neonatal asphyxia risk prediction models have been reported at home and abroad. However, most of these models are based on single-center studies, with small sample size, and some studies only use chi-square test for single factor analysis without further multi-factor analysis to determine the independent risk factors of neonatal asphyxia. In addition, due to the differences in medical conditions and race, the prediction models reported abroad cannot be directly applied to the prediction of Chinese population.

[0004] Therefore, how to realize early and accurate identification of high-risk neonates is a problem to be solved by those skilled in the art. SUMMARY

[0005] The embodiments of the present application provide a method and device for constructing a neonatal asphyxia risk prediction model, which can realize early and accurate identification of high-risk neonates.

[0006] The first aspect of the present application provides a method for constructing a neonatal asphyxia risk prediction model, comprising: obtaining clinical feature data of a neonatal asphyxia group and a healthy neonatal group, the clinical feature data including maternal information data, fetal information data and other information data; performing single factor analysis on clinical risk factors between the neonatal asphyxia group and the healthy neonatal group, and screening single factor prediction variables related to neonatal asphyxia; performing binary Logistic regression analysis on the screened single factor prediction variables as independent variables and whether asphyxia occurs when the neonate is born as dependent variable, and establishing a neonatal asphyxia risk prediction model; wherein the binary Logistic regression analysis uses forward stepwise method.

[0007] Optionally, maternal information data may include any one or more of the following: age, hypertension, diabetes, thyroid disease, intrahepatic cholestasis, uterine fibroids, thrombocytopenia, multiple pregnancy, scarred uterus, primiparous, first-time mother, and antepartum fever. Fetal information data includes any one or more of the following: premature infant, macrosomic infant, low birth weight infant, sex, fetal distress, fetal growth restriction, and fetal congenital diseases; Other information includes any one or more of the following: umbilical cord abnormalities, placental abnormalities, amniotic fluid abnormalities, fetal position abnormalities, premature rupture of membranes, mode of pregnancy, and mode of delivery.

[0008] Optionally, prior to the binary logistic regression analysis, the method may also include: Multicollinearity diagnosis was performed on univariate predictor variables, which included hypertension, multiple pregnancy, preterm infant, low birth weight infant, sex, fetal distress, fetal growth restriction, placental abnormalities, amniotic fluid abnormalities, abnormal fetal position, and mode of delivery.

[0009] Optionally, the selected univariate predictor variables are used as independent variables, and whether asphyxia occurred at birth is used as the dependent variable, for binary logistic regression analysis, including: Risk factors selected for constructing a neonatal asphyxia risk prediction model were identified using a forward stepwise screening method. These risk factors included premature birth, low birth weight, sex, fetal distress, placental abnormalities, amniotic fluid abnormalities, and abnormal fetal position.

[0010] Optionally, the neonatal asphyxia risk prediction model is a logistic regression equation constructed based on the partial regression coefficients of seven risk factors: premature infants, low birth weight infants, sex, fetal distress, placental abnormalities, amniotic fluid abnormalities, and fetal position abnormalities.

[0011] Optionally, after establishing a neonatal asphyxia risk prediction model, the method further includes: Based on the neonatal asphyxia risk prediction model, a nomogram is plotted. The nomogram is used to map the values ​​of multiple risk factors to corresponding scores, and the probability of neonatal asphyxia risk is visualized by summing the scores of each risk factor.

[0012] Optionally, after drawing the nomogram, the method further includes: ROC curves were used to evaluate the discrimination of the neonatal asphyxia risk prediction model, and calibration curves and Hosmer-Lemeshow goodness-of-fit tests were used to evaluate the calibration of the neonatal asphyxia risk prediction model.

[0013] The second aspect of this application provides a device for constructing a neonatal asphyxia risk prediction model, comprising: The data acquisition module is used to acquire clinical characteristic data of the neonatal asphyxia group and the healthy neonatal group. The clinical characteristic data includes maternal information data, fetal information data and other information data. The univariate analysis module is used to perform univariate analysis of clinical risk factors between the neonatal asphyxia group and the healthy neonatal group, and to screen out univariate predictive variables related to neonatal asphyxia. The regression modeling module is used to perform binary logistic regression analysis on the selected single-factor predictor variables as independent variables and whether the newborn experiences asphyxia at birth as the dependent variable, to establish a newborn asphyxia risk prediction model; the binary logistic regression analysis adopts the forward stepwise method.

[0014] A third aspect of this application provides an electronic device, comprising: One or more processors; A memory on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the neonatal asphyxia risk prediction model construction method as described in any of the above.

[0015] The fourth aspect of this application provides a computer storage medium for storing a program, which, when executed, is used to implement the neonatal asphyxia risk prediction model construction method as described in any of the preceding claims.

[0016] Compared with the prior art, the beneficial effects of the present invention include: The neonatal asphyxia risk prediction model constructed in this invention is more scientific and rigorous, exhibiting excellent predictive performance. Based on multi-center, large-sample clinical data, this invention employs a modeling process progressing from univariate analysis to multivariate binary logistic regression stepwise forward method, effectively screening out truly independent risk factors. Compared to some existing technologies that only use univariate analysis, the prediction model constructed in this invention has higher discriminative power and better calibration, ensuring the accuracy and reliability of the prediction.

[0017] This invention clarifies a key combination of independent risk factors applicable to the Chinese population. Through the aforementioned scientific statistical analysis methods, this invention, for the first time, systematically identifies and confirms seven specific perinatal factors—prematurity, low birth weight, sex, fetal distress, placental abnormalities, amniotic fluid abnormalities, and abnormal fetal position—as independent risk factors for predicting neonatal asphyxia. This provides crucial evidence for accurate risk assessment of neonatal asphyxia in the Chinese clinical context, overcoming the problem that some foreign models are inapplicable due to population differences.

[0018] This invention has strong clinical applicability, facilitating rapid assessment and decision support. Furthermore, it transforms the predictive model into an intuitive visual nomogram. This tool allows clinicians to quickly obtain individualized asphyxia risk probabilities based on the specific circumstances of the pregnant woman and fetus, without complex calculations. This not only helps in the early and efficient identification of high-risk newborns before and during delivery, providing clear guidance for timely targeted interventions and resuscitation preparation, but also significantly improves the applicability and operability of the model in medical institutions at all levels, especially in resource-constrained environments.

[0019] This invention helps improve neonatal health outcomes. By applying the predictive model or nomogram of this invention, effective early warning of neonatal asphyxia risk can be achieved, thereby promoting early intervention, early prevention, and individualized management. This is expected to reduce the incidence of neonatal asphyxia and its related complications, and has significant clinical and social value in improving neonatal survival rates and long-term prognosis. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this application 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 recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 A flowchart illustrating a method for constructing a neonatal asphyxia risk prediction model, provided in an embodiment of this application; Figure 2 A nodal diagram of a neonatal asphyxia risk prediction model provided in this application embodiment; Figure 3 A schematic diagram of the ROC curve of a neonatal asphyxia risk prediction model provided in this application embodiment; Figure 4 This is a schematic diagram of a calibration curve for a neonatal asphyxia risk prediction model provided in an embodiment of this application. Figure 5 A schematic diagram of a neonatal asphyxia risk prediction model construction device provided in this application embodiment; Figure 6 This is a schematic diagram of a neonatal asphyxia risk prediction model construction device provided in an embodiment of this application. Detailed Implementation

[0022] This application provides a method and apparatus for constructing a neonatal asphyxia risk prediction model, which can achieve early and accurate identification of high-risk newborns.

[0023] See Figure 1This figure is a flowchart illustrating a method for constructing a neonatal asphyxia risk prediction model according to an embodiment of this application. The method for constructing a neonatal asphyxia risk prediction model provided in this embodiment can be implemented, for example, through the following steps S101-103.

[0024] S101: Obtain clinical characteristic data for the neonatal asphyxia group and the healthy neonatal group.

[0025] In this embodiment, the clinical characteristic data includes maternal information data, fetal information data, and other information data. Maternal information data includes any one or more of the following: age, hypertension, diabetes, thyroid disease, intrahepatic cholestasis, uterine fibroids, thrombocytopenia, multiple pregnancy, scarred uterus, primiparous woman, primiparous woman, and antepartum fever; fetal information data includes any one or more of the following: premature infant, macrosomia, low birth weight infant, sex, fetal distress, fetal growth restriction, and congenital fetal diseases; other information data includes any one or more of the following: umbilical cord abnormalities, placental abnormalities, amniotic fluid abnormalities, abnormal fetal position, premature rupture of membranes, pregnancy mode, and delivery mode.

[0026] Specifically, the data in this embodiment of the invention comes from three hospitals in Xiamen City, and relevant clinical characteristic data are retrieved through the electronic medical record systems of these three hospitals. The inclusion criteria are newborns whose outcomes at discharge are cured or improved, and whose mothers have complete prenatal, intrapartum, and postpartum clinical data. The exclusion criteria are those who discontinued treatment during hospitalization, died, or whose mothers had incomplete clinical data. The final dataset consists of 600 individuals, including 290 cases in the neonatal asphyxia group and 310 cases in the healthy newborn group.

[0027] Clinical characteristic data collection includes the following three aspects: (1) Maternal information data: age, hypertension, diabetes, thyroid disease, intrahepatic cholestasis, uterine fibroids, thrombocytopenia, multiple pregnancy, scarred uterus, primiparous women, primiparous women and antepartum fever.

[0028] (2) Fetal information data: premature infants, macrosomic infants, low birth weight infants, sex, fetal distress, fetal growth restriction and fetal congenital diseases.

[0029] (3) Other information data: umbilical cord abnormalities, placental abnormalities, amniotic fluid abnormalities, fetal position abnormalities, premature rupture of membranes, pregnancy mode and delivery mode.

[0030] S102: Univariate analysis of clinical risk factors between the neonatal asphyxia group and the healthy neonatal group was conducted to screen out univariate predictive variables associated with neonatal asphyxia.

[0031] In the embodiments of this application, the univariate predictive variables include hypertension, multiple pregnancy, premature infant, low birth weight infant, sex, fetal distress, fetal growth restriction, placental abnormality, amniotic fluid abnormality, abnormal fetal position, and mode of delivery.

[0032] Specifically, the clinical characteristic data were statistically analyzed using the Statistical Package for the Social Sciences (SPSS). Count data were expressed as frequencies and percentages, and the chi-square test was used for inter-group comparisons, with p < 0.05 considered statistically significant. In this embodiment of the invention, the clinical risk factors between the neonatal asphyxia group and the healthy neonatal group were first statistically analyzed to preliminarily screen out univariate predictive variables related to neonatal asphyxia. As shown in Table 1, Table 1 compares the clinical risk factors between the neonatal asphyxia group and the healthy neonatal group provided in this embodiment of the application. The results show that there are significant differences between the two groups in terms of hypertension, multiple pregnancy, preterm infants, low birth weight infants, sex, fetal distress, fetal growth restriction, placental abnormalities, amniotic fluid abnormalities, abnormal fetal position, and delivery method.

[0033] Table 1 Comparison of clinical risk factors between neonatal asphyxia group and healthy neonate group.

[0034] S103: Using the selected univariate predictor variables as independent variables and whether the newborn experiences asphyxia at birth as the dependent variable, perform binary logistic regression analysis to establish a newborn asphyxia risk prediction model.

[0035] In this embodiment, multicollinearity diagnosis is first performed on the univariate predictor variables. Then, a forward stepwise method is used to screen for risk factors used to construct a neonatal asphyxia risk prediction model. These risk factors include preterm birth, low birth weight, sex, fetal distress, placental abnormalities, amniotic fluid abnormalities, and abnormal fetal position. The neonatal asphyxia risk prediction model is a logistic regression equation constructed based on the partial regression coefficients of the seven risk factors: preterm birth, low birth weight, sex, fetal distress, placental abnormalities, amniotic fluid abnormalities, and abnormal fetal position.

[0036] Specifically, the univariate predictor variables obtained from the above univariate analysis were used as independent variables, and whether neonatal asphyxia occurred at birth was used as the dependent variable for binary logistic regression analysis. First, tolerance or variance inflation factor was used to determine if multicollinearity existed among the independent variables, and then multivariate analysis was performed using the forward stepwise method. A neonatal asphyxia risk prediction model was established based on the partial regression coefficients β of the clinical risk factors obtained from the binary logistic regression analysis. Tolerance or variance inflation factor was used to diagnose whether multicollinearity existed among the independent variables; a tolerance less than 0.1 or a variance inflation factor greater than 10 indicated the presence of multicollinearity. In this embodiment, after collinearity statistics, the tolerance range of the 11 independent variables was 0.320-0.976, all greater than 0.1, and the variance inflation factor ranged from 1.025-3.128, all less than 10. These results indicate that there was no multicollinearity among the 11 independent variables.

[0037] Refer to Table 2, which presents the results of a binary logistic regression analysis of neonatal asphyxia provided in the embodiments of this application. In the table, β is the partial regression coefficient; SE is the standard error; Wald is the chi-square value; OR is the odds ratio; and CI is the confidence interval. The binary logistic regression analysis results show that preterm birth, low birth weight, sex, fetal distress, placental abnormalities, amniotic fluid abnormalities, and abnormal fetal position are independent risk factors for neonatal asphyxia.

[0038] Table 2. Results of binary logistic regression analysis for neonatal asphyxia

[0039] A neonatal asphyxia risk prediction model was established based on the partial regression coefficients β of clinical risk factors obtained from binary logistic regression analysis, namely, Logit(p) = -1.857 + 1.753 × preterm infant (No: 0, Yes: 1) + 1.061 × low birth weight infant (No: 0, Yes: 1) + 0.684 × sex (Female: 0, Male: 1) + 2.523 × fetal distress (No: 0, Yes: 1) + 0.903 × placental abnormality (No: 0, Yes: 1) + 0.927 × amniotic fluid abnormality (No: 0, Yes: 1) + 0.907 × abnormal fetal position (No: 0, Yes: 1). Where p is the predicted probability of the risk prediction model; the larger the Logit(p), the higher the risk of neonatal asphyxia.

[0040] In one implementation of this application, a nomogram is drawn based on a neonatal asphyxia risk prediction model; wherein, the nomogram is used to map the values ​​of multiple risk factors to corresponding scores, and to visualize and predict the risk probability of neonatal asphyxia by summing the scores of each risk factor.

[0041] Specifically, Figure 2This application provides a nomogram of a neonatal asphyxia risk prediction model. To provide a more intuitive understanding of the probability of neonatal asphyxia at birth, the model is visualized and plotted. Figure 2 The nomogram shown is illustrated. Each clinical risk factor corresponds to a score at the top of the nomogram. The sum of the scores for all known risk factors is then compared to the risk probability value at the bottom of the nomogram to represent the probability of the newborn experiencing asphyxia at birth. For example, if a pregnant woman's prenatal checkup revealed low fetal weight and placental abnormalities, her known risk factor score would be approximately 78 points. This would indicate a newborn asphyxia risk of 0.5-0.6, meaning the probability of asphyxia at birth is between 50% and 60%. If premature birth occurs, the probability increases to approximately 85%, classifying the pregnant woman as high-risk. Close monitoring of labor is crucial, and preparations should be made for neonatal resuscitation at any time. Converting the predictive model into a nomogram makes it simpler, more intuitive, and easier for clinical application.

[0042] In one implementation of this application, ROC curves are used to evaluate the discrimination of the neonatal asphyxia risk prediction model, and calibration curves and Hosmer-Lemeshow goodness-of-fit tests are used to evaluate the calibration of the neonatal asphyxia risk prediction model.

[0043] Specifically, the area under the ROC curve (AUC) is used to evaluate the discrimination ability of the prediction model. A larger AUC indicates better discrimination ability; AUC < 0.6 indicates poor discrimination ability; AUC between 0.6 and 0.75 indicates some discrimination ability; and AUC > 0.75 indicates good discrimination ability. The calibration curve and Hosmer-Lemeshow goodness-of-fit test are used to evaluate the calibration of the prediction model. A closer fit between the calibration curve and the ideal curve, and a p > 0.05 in the Hosmer-Lemeshow test results, indicate better calibration.

[0044] In embodiments of the present invention, such as Figure 3 As shown, Figure 3 This is a schematic diagram of the ROC curve for a neonatal asphyxia risk prediction model provided in an embodiment of this application. The AUC of the ROC curve is 0.844 (95% CI: 0.813-0.875), the sensitivity is 0.679, and the specificity is 0.874, indicating that the prediction model has good discriminative power. Figure 4 As shown, Figure 4This is a schematic diagram of the calibration curve for a neonatal asphyxia risk prediction model provided in this application embodiment. The calibration curve was obtained by repeatedly sampling 1000 times using the bootstrap method in repeated sampling. The results show that the calibrated prediction curve is close to the ideal curve, and the χ² value in the Hosmer-Lemeshow goodness-of-fit test is also good. 2 =9.828, p=0.132, indicating that the prediction model has good calibration. These results demonstrate that the neonatal asphyxia risk prediction model constructed in this invention has good predictive efficacy.

[0045] Beneficial effects: This invention utilizes a case-control study with a large sample size of clinically relevant data from newborns at birth in multiple centers. Based on independent risk factors for neonatal asphyxia, a neonatal asphyxia risk prediction model is constructed, and the model is visualized by plotting a nomogram. Finally, the prediction model is evaluated in terms of discrimination and calibration. This is of great significance for early identification of high-risk groups and timely development of individualized interventions to prevent neonatal asphyxia and improve neonatal survival rates in clinical practice.

[0046] Based on the methods provided in the above embodiments, this application also provides a neonatal asphyxia risk prediction model construction device, which is described below with reference to the accompanying drawings.

[0047] See Figure 5 The figure is a schematic diagram of a neonatal asphyxia risk prediction model construction device provided in an embodiment of this application.

[0048] The neonatal asphyxia risk prediction model construction device 500 provided in this application embodiment includes: a data acquisition module 501, a single-factor analysis module 502, and a regression modeling module 503.

[0049] The data acquisition module 501 is used to acquire clinical characteristic data of the neonatal asphyxia group and the healthy neonatal group. The clinical characteristic data includes maternal information data, fetal information data and other information data. The univariate analysis module 502 is used to perform univariate analysis on clinical risk factors between the neonatal asphyxia group and the healthy neonatal group, and to screen out univariate predictive variables related to neonatal asphyxia. The regression modeling module 503 is used to perform binary logistic regression analysis on the selected single-factor predictor variables as independent variables and whether the newborn experiences asphyxia at birth as the dependent variable, to establish a newborn asphyxia risk prediction model; the binary logistic regression analysis adopts the forward stepwise method.

[0050] In one possible implementation, the maternal information data includes any one or more of the following: age, hypertension, diabetes, thyroid disease, intrahepatic cholestasis, uterine fibroids, thrombocytopenia, multiple pregnancy, scarred uterus, primiparous, first-time mother, and antepartum fever. Fetal information data includes any one or more of the following: premature infant, macrosomic infant, low birth weight infant, sex, fetal distress, fetal growth restriction, and fetal congenital diseases; Other information includes any one or more of the following: umbilical cord abnormalities, placental abnormalities, amniotic fluid abnormalities, fetal position abnormalities, premature rupture of membranes, mode of pregnancy, and mode of delivery.

[0051] In one possible implementation, the neonatal asphyxia risk prediction model building device 500 further includes a multicollinearity diagnostic module, specifically used for: Multicollinearity diagnosis was performed on univariate predictor variables, which included hypertension, multiple pregnancy, preterm infant, low birth weight infant, sex, fetal distress, fetal growth restriction, placental abnormalities, amniotic fluid abnormalities, abnormal fetal position, and mode of delivery.

[0052] In one possible implementation, the regression modeling module 503 is specifically used for: Risk factors selected for constructing a neonatal asphyxia risk prediction model were identified using a forward stepwise screening method. These risk factors included premature birth, low birth weight, sex, fetal distress, placental abnormalities, amniotic fluid abnormalities, and abnormal fetal position.

[0053] In one possible implementation, the neonatal asphyxia risk prediction model is a logistic regression equation constructed based on the partial regression coefficients of seven risk factors: preterm infants, low birth weight infants, sex, fetal distress, placental abnormalities, amniotic fluid abnormalities, and fetal position abnormalities.

[0054] In one possible implementation, the neonatal asphyxia risk prediction model building device 500 further includes a drawing module, specifically used for: Based on the neonatal asphyxia risk prediction model, a nomogram is plotted. The nomogram is used to map the values ​​of multiple risk factors to corresponding scores, and the probability of neonatal asphyxia risk is visualized by summing the scores of each risk factor.

[0055] In one possible implementation, the neonatal asphyxia risk prediction model building device 500 further includes an evaluation module, specifically used for: ROC curves were used to evaluate the discrimination of the neonatal asphyxia risk prediction model, and calibration curves and Hosmer-Lemeshow goodness-of-fit tests were used to evaluate the calibration of the neonatal asphyxia risk prediction model.

[0056] Since the neonatal asphyxia risk prediction model construction device 500 is a device corresponding to the neonatal asphyxia risk prediction model construction method provided in the above method embodiments, the specific implementation of each unit of the neonatal asphyxia risk prediction model construction device 500 is based on the same concept as in the above method embodiments. Therefore, for the specific implementation of each unit of the neonatal asphyxia risk prediction model construction device 500, please refer to the description of the neonatal asphyxia risk prediction model construction method in the above method embodiments, and it will not be repeated here.

[0057] This application embodiment also provides a neonatal asphyxia risk prediction model construction device, the device including: a processor and a memory; The memory is used to store instructions; The processor is used to execute the instructions in the memory to perform the neonatal asphyxia risk prediction model construction method mentioned in the above embodiments.

[0058] It should be noted that the hardware structure of the neonatal asphyxia risk prediction model construction device provided in this application embodiment can be as follows: Figure 6 The structure shown, Figure 6 This is a schematic diagram of a neonatal asphyxia risk prediction model construction device provided in an embodiment of this application.

[0059] Please see Figure 6 As shown, the neonatal asphyxia risk prediction model building device 600 includes: a processor 610, a communication interface 620, and a memory 630. The number of processors 610 in the neonatal asphyxia risk prediction model building device 600 can be one or more. Figure 6 Taking a processor as an example. In this embodiment, the processor 610, communication interface 620, and memory 630 can be connected via a bus system or other means. Figure 6 Taking the connection between China and Israel via the Bus System 640 as an example.

[0060] Processor 610 may be a central processing unit (CPU), a network processor (NP), or a combination of a CPU and an NP. Processor 610 may further include hardware chips. These hardware chips may be application-specific integrated circuits (ASICs), programmable logic devices (PLDs), or combinations thereof. The PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), generic array logic (GAL), or any combination thereof.

[0061] The memory 630 may include volatile memory, such as random-access memory (RAM); the memory 630 may also include non-volatile memory, such as flash memory, hard disk drive (HDD) or solid-state drive (SSD); the memory 630 may also include a combination of the above types of memory.

[0062] Optionally, the memory 630 stores an operating system and programs, executable modules, or data structures, or subsets thereof, or extended sets thereof. The programs may include various operation instructions for implementing various operations. The operating system may include various system programs for implementing various basic business processes and handling hardware-based tasks. The processor 610 can read the programs in the memory 630 to implement the neonatal asphyxia risk prediction model construction method provided in this application embodiment.

[0063] The bus system 640 can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The bus system 640 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 6 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0064] This application also provides a computer-readable storage medium including instructions that, when executed on a computer, cause the computer to perform the neonatal asphyxia risk prediction model construction method mentioned in the above embodiments.

[0065] This application also provides a computer program product containing instructions that, when run on a computer, cause the computer to execute the neonatal asphyxia risk prediction model construction method mentioned in the above embodiments. Although the invention has been specifically shown and described in conjunction with preferred embodiments, those skilled in the art should understand that various changes in form and detail may be made to the invention without departing from the spirit and scope of the invention as defined in the appended claims, all of which shall be within the scope of protection of the invention.

Claims

1. A method for constructing a neonatal asphyxia risk prediction model, characterized in that, include: Acquire clinical characteristic data of neonatal asphyxia group and healthy neonatal group, the clinical characteristic data including maternal information data, fetal information data and other information data; Univariate analysis was performed on the clinical risk factors between the neonatal asphyxia group and the healthy neonatal group to screen out univariate predictive variables related to neonatal asphyxia. The selected univariate predictor variables were used as independent variables, and whether the newborn experienced asphyxia at birth was used as the dependent variable. A binary logistic regression analysis was performed to establish a newborn asphyxia risk prediction model. The binary logistic regression analysis used the forward stepwise method.

2. The method for constructing a neonatal asphyxia risk prediction model according to claim 1, characterized in that, The maternal information data includes any one or more of the following: age, hypertension, diabetes, thyroid disease, intrahepatic cholestasis, uterine fibroids, thrombocytopenia, multiple pregnancy, scarred uterus, primiparous woman, primiparous woman, and antepartum fever. The fetal information data includes any one or more of the following: premature infant, macrosomic infant, low birth weight infant, sex, fetal distress, fetal growth restriction, and fetal congenital diseases. The other information data includes any one or more of the following: umbilical cord abnormalities, placental abnormalities, amniotic fluid abnormalities, fetal position abnormalities, premature rupture of membranes, mode of pregnancy, and mode of delivery.

3. The method for constructing a neonatal asphyxia risk prediction model according to claim 1 or 2, characterized in that, Prior to the binary logistic regression analysis, the method further includes: Multicollinearity diagnosis was performed on the univariate predictor variables, which included hypertension, multiple pregnancy, premature infant, low birth weight infant, sex, fetal distress, fetal growth restriction, placental abnormalities, amniotic fluid abnormalities, abnormal fetal position, and mode of delivery.

4. The method for constructing a neonatal asphyxia risk prediction model according to claim 1, characterized in that, The step involves using the selected univariate predictor variables as independent variables and whether asphyxia occurred at birth as the dependent variable, and performing a binary logistic regression analysis, including: The risk factors selected by the forward stepwise method for constructing the neonatal asphyxia risk prediction model include premature infants, low birth weight infants, sex, fetal distress, placental abnormalities, amniotic fluid abnormalities, and abnormal fetal position.

5. The method for constructing a neonatal asphyxia risk prediction model according to claim 4, characterized in that, The neonatal asphyxia risk prediction model is a logistic regression equation constructed based on the partial regression coefficients of seven risk factors: premature infants, low birth weight infants, sex, fetal distress, placental abnormalities, amniotic fluid abnormalities, and fetal position abnormalities.

6. The method for constructing a neonatal asphyxia risk prediction model according to claim 1, characterized in that, After establishing the neonatal asphyxia risk prediction model, the method further includes: Based on the neonatal asphyxia risk prediction model, a nomogram is plotted to visualize the neonatal asphyxia risk prediction model. The nomogram is used to map the values ​​of multiple risk factors to corresponding scores, and the probability of neonatal asphyxia risk is visualized by summing the scores of each risk factor.

7. The method for constructing a neonatal asphyxia risk prediction model according to claim 6, characterized in that, After drawing the column chart, the method further includes: The discrimination of the neonatal asphyxia risk prediction model was evaluated using receiver operating characteristic (ROC) curves, and the calibration of the neonatal asphyxia risk prediction model was evaluated using calibration curves and the Hosmer-Lemeshow goodness-of-fit test.

8. A device for constructing a neonatal asphyxia risk prediction model, characterized in that, include: The data acquisition module is used to acquire clinical characteristic data of the neonatal asphyxia group and the healthy neonatal group. The clinical characteristic data includes maternal information data, fetal information data and other information data. The univariate analysis module is used to perform univariate analysis on clinical risk factors between the neonatal asphyxia group and the healthy neonatal group, and to screen out univariate predictive variables related to neonatal asphyxia. The regression modeling module is used to perform binary logistic regression analysis on the selected single-factor predictor variables as independent variables and whether asphyxia occurs at birth as the dependent variable, to establish a neonatal asphyxia risk prediction model; wherein the binary logistic regression analysis adopts the forward stepwise method.

9. An electronic device, characterized in that, The device includes: a processor and a memory; The memory is used to store instructions; The processor is configured to execute the instructions in the memory to perform the method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, Including instructions that, when run on a computer, cause the computer to perform the method described in any one of claims 1-7 above.

Images