基于机器学习算法构建初产妇合并巨大儿试产结局的预测模型  

Constructing the prediction model of labor outcomes of nulliparous women combined with macrosomia based on machine learning algorithms

作　　者：邓晨晨 余涛 陈红波[1] Deng Chenchen;Yu Tao;Chen Hongbo(Anhui Women and Children's Medical Center,Hefei Maternal and Child Health Hospital,Hefei230000;Anhui Mental Health Center,Hefei Fourth People's Hospital,Hefei230000)

机构地区：[1]安徽省妇女儿童医学中心、合肥市妇幼保健院,合肥230000 [2]安徽省精神卫生中心合肥市第四人民医院,合肥230000

出　　处：《现代妇产科进展》2024年第9期662-665,共4页Progress in Obstetrics and Gynecology

基　　金：安徽省重点研究与开发计划-临床医学研究转化专项(No:202204295107020050);安徽省首届“青年江淮名医”培养项目(No:2022-392)。

摘　　要：目的:探究妊娠合并巨大儿的初产妇经阴道试产结局的影响因素及预测模型。方法:收集2022年1月至2023年12月于安徽省妇女儿童医学中心住院并经阴道试产的巨大儿初产妇临床资料,根据分娩结局分为阴道分娩组及剖宫产组。单因素分析比较两组间资料,LASSO分析进一步筛选变量,分别采用支持向量机(SVM)、一般线性模型(GLM)、K最近邻(KNN)、随机森林(RF)建立预测模型,最后得出最佳模型。结果:阴道分娩组产妇的年龄、孕前及产前体质量指数(BMI)低于剖宫产组,阴道分娩组产妇身高、Bishop评分、胎儿股骨长高于剖宫产组,阴道分娩组自然临产比例高于剖宫产组,差异均有统计学意义。LASSO分析筛选出变量,结合机器学习算法构建预测模型,最终通过受试者工作曲线(ROC)对预测模型进行比较,GLM在4种预测模型中表现最佳,曲线下面积(AUC)达0.7699。结论:妊娠合并巨大儿的初产妇人群,可运用产妇年龄、身高、孕前及产前BMI、Bishop评分、胎儿股骨长、临产方式结合机器学习算法构建GLM预测模型,为初产妇合并巨大儿分娩方式的选择提供参考依据。Objective:To explore the influencing factors and prediction models of vaginal trial of labor outcomes in nulliparous women combined with macrosomia.Methods:To collect clinical data of nulliparous women combined with macrosomia performed vaginal trial of labor from January 2022 to December 2023 in Anhui Women and Children's Medical Center.According to labor outcomes,they were divided into two groups,including vaginal labor group and caesarean section group.Univariate analysis was conducted between two groups,and LASSO analysis further screened the variables.Support vector machine(SVM),general linear model(GLM),K-nearest neighbor(KNN),and random forest(RF)were used to establish predictive models,respectively.The optimal model was ultimately determined.Results:Maternal age,pre-pregnancy and prenatal body mass index(BMI)in the vaginal labor group were lower than those in the cesarean section group.Maternal height,Bishop score,and fetal femur length in the vaginal labor group were higher than those in the cesarean section group.The proportion of natural labor in the vaginal labor group was higher than that in the cesarean section group.The differences were statistically significant.The above variables were regressed by LASSO analysis and combined with machine learning algorithms to construct prediction models.Finally,all prediction models were comparable by receiver operating characteristic(ROC),and GLM performed the best among the four prediction models,yielding an area under the curve(AUC)of 0.7699.Conclusion:For the population of nulliparous women combined with macrosomia,the prediction model developed using the machine learning method and maternal age,height,pre-pregnancy and prenatal BMI,Bishop score,fetal femur length,and labor mode is able to provide a reference basis for the selection of delivery method for nulliparous women combined with macrosomia.

关 键 词：巨大儿 初产妇 试产 影响因素 机器学习 预测模型 

分 类 号：R714[医药卫生—妇产科学]