[1]于文凯,孙梦怡,李仁哲,等.基于可解释机器学习算法和网页计算器构建孕晚期B族链球菌感染不良妊娠结局风险的预测模型[J].现代检验医学杂志,2026,41(02):37-45.[doi:10.3969/j.issn.1671-7414.2026.02.007]
　YU Wenkai,SUN Mengyi,LI Renzhe,et al.Development of a Risk Predictive Model for Adverse Pregnancy Outcomes Associated with Group B Streptococcus Infection in Late Pregnancy Based on Interpretable Machine Learning Methods and a Web-based Calculator[J].Journal of Modern Laboratory Medicine,2026,41(02):37-45.[doi:10.3969/j.issn.1671-7414.2026.02.007]
点击复制

基于可解释机器学习算法和网页计算器构建孕晚期B族链球菌感染不良妊娠结局风险的预测模型()

《现代检验医学杂志》[ISSN:/CN:]

卷:
第41卷
期数:
2026年02期
页码:
37-45
栏目:
论著
出版日期:
2026-03-15

文章信息/Info

Title:
Development of a Risk Predictive Model for Adverse Pregnancy Outcomes Associated with Group B Streptococcus Infection in Late Pregnancy Based on Interpretable Machine Learning Methods and a Web-based Calculator
文章编号:
1671-7414（2026）02-037-09
作者:
于文凯，孙梦怡，李仁哲，郑 康
济宁市第一人民医院医学检验科，山东济宁 272000
Author(s):
YU Wenkai，SUN Mengyi，LI Renzhe，ZHENG Kang
Department of Clinical Laboratory，Jining No.1 People’s Hos-pital，Shandong Jining 272000，China
关键词:
机器学习; 网页计算器; B族链球菌; 妊娠结局; 预测模型
分类号:
R378.12；R446
DOI:
10.3969/j.issn.1671-7414.2026.02.007
文献标志码:
A
摘要:
目的?利用机器学习(ML)算法和网页计算器构建孕晚期B族链球菌(GBS)感染不良妊娠结局风险的预测模型,探索结局相关因子。方法回顾性分析2022年4月~2025年4月期间,济宁市第一人民医院GBS筛查为阳性的625例孕晚期产妇的病历资料。收集患者临床资料、实验室检查数据、追踪妊娠结局,采用最小绝对收缩和选择算子(LASSO)算法筛选关键特征。运用决策树(DT)、随机森林(RF)、极端梯度提升树(XGBoost)、支持向量机(SVM)、K最近邻(KNN)和逻辑回归(LR)6种ML算法建立预测模型,并筛选出最优模型,进一步采用沙普利加和解释(SHAP)算法对最佳模型进行解释性分析。开发一种网页计算器,能够在线预测不良妊娠结局风险。结果6种ML算法模型中,LR模型展现出最佳的预测能力与稳定性,曲线下面积(AUC)为0.803。SHAP蜂群图和特征重要性图显示,孕周、体重指数(BMI)、全身炎症反应指数(SIRI)、年龄、单核细胞与淋巴细胞计数比值(MLR)和妊娠期糖尿病(GDM)是GBS感染孕妇发生不良妊娠结局关键因素。开发的网页计算器(https://adverse-pregnancy-outcomes.shinyapps.io/adverse-pregnancy-outcomes/)初步实现了预测模型的可视化应用。结论结合临床和实验室数据构建的ML模型及网页计算器,能够有效评估GBS感染孕妇发生不良妊娠结局的风险,具有较高地准确度。
Abstract:
Objective To develop a predictive model for the risk of adverse pregnancy outcome associated with late pregnancy Group B Streptococcus infection (GBS) based on interpretable machine learning (ML) algorithms and an online calculator to explore outcome-related factors. Methods A retrospective analysis was performed on the medical records of 625 women in late pregnancy with positive GBS screening results at Jining No.1 People’s Hospital between April 2022 to April 2025. Clinical data, laboratory indices, and pregnancy outcomes were collected. Feature selection was subsequently performed using the LASSO algorithm. Six machine learning algorithms were employed for model development: Decision Tree (DT), Random Forest (RF), Extreme G radient Boosting (XGBoost), Support Vector Machine (SVM), K-Nearest Neighbor (KNN), and Logistic Regression (LR). The optimal model was selected, and subjected to interpretative analysis using the Shapley Additive Explanations (SHAP) algorithm. A web-based calculator was developed to predict the risk of adverse pregnancy outcomes. Results Among the six machine learning mod-els, the Logistic Regression (LR) model showed the best predictive performance and stability, with an area under the curve (AUC) of 0.803. SHAP swarm plots and feature importance plots indicated that gestational age, body mass index, systemic inflammation response index, age, the monocyte-to-lymphocyte ratio, and gestational diabetes were key factors for the occurrence of adverse pregnancy outcomes in GBS-infected pregnant women. The web calculator (https://adverse-pregnancy-outcomes.shinyapps.io/ad-verse-pregnancy-outcomes/) enables preliminary visualization of the predictive model. Conclusions The machine learning model and web-based calculator constructed using clinical information and laboratory indicators can effectively evaluate the risk of adverse pregnancy outcomes in GBS-infected pregnant women with high accuracy.

参考文献/References:

[1] LITMAN E, YOUNG B, SPIEL M. Novel insights on group B Streptococcus in pregnancy[J]. Clinical Ob-stetrics and Gynecology, 2024, 67(3): 633-643.
[2] 杨柳,付欣,张小飞.胎膜早破孕妇生殖道无乳链球菌血清型分型和耐药基因分析[J].现代检验医学杂志, 2023, 38(3): 123-127. YANG L, FU X, ZHANG X F. Serotyping and resistance gene analysis of reproductive Streptococcus agalactiae in women with premature rupture of membranes[J]. Journal of Modern Laboratory Medicine, 2023, 38(3): 123-127.
[3] JIN G X, LIU L H, WANG X L, et al. Maternal and neonatal outcomes of Group B Streptococcus colonization: a retro-spective study[J]. BMC Infectious Diseases, 2025, 25(1): 94.
[4] GON?ALVES B P, PROCTER S R, PAUL P, et al. Group B Streptococcus infection during pregnancy and infancy: estimates of regional and global burden [J]. The Lancet Global Health, 2022, 10(6): e807 e819.
[5] LEE K W, YAP S F, MURDAN S, et al. Maternal and neonatal group B Streptococcus colonisation: a sys-tematic review and the meta-analysis of matched-pair studies[J]. Acta Paediatrica, 2024, 113(5): 892-911.
[6] DELARA M, VADLAMUDI N K, SADARANGANI M. Strategies to prevent early and late-onset group B Streptococcal infection via interventions in pregnan-cy[J]. Pathogens, 2023, 12(2): 229.
[7] MEI J Y, SILVERMAN N S. Group B Streptococcus in pregnancy[J]. Obstetrics and Gynecology Clinics of North America, 2023, 50(2): 375-387.
[8] WU M Q, DENG Y B, WANG X Y, et al. Development of risk prediction nomogram for neonatal sepsis in group B Streptococcus-colonized mothers: a retrospec-tive study[J]. Scientific Reports, 2024, 14(1): 5629.
[9] 郝晓燕,周磊,白丛霞,等.基于机器学习方法建立血清多种肿瘤标志物联合区分胃炎和胃癌的三种模型及评价[J].现代检验医学杂志, 2023, 38(2): 124-128. HAO X Y, ZHOU L, BAI C X, et al. Establishment and evaluation of three models for the distinguish gastritis and gastric cancer with multiple serum tumor markers based on machine learning methods[J]. Journal of Mod-ern Laboratory Medicine, 2023, 38(2): 124-128.
[10] 张乔盟,杨蕊菲,冯飞雪,等.原发性肝细胞癌相关风险评估模型的最新研究进展[J].现代检验医学杂志, 2025, 40(1): 203-207. ZHANG Q M, YANG R F, FENG F X, et al. Latest research progress on risk scoring models for primary hepatocellular carcinoma[J]. Journal of Modern Labo-ratory Medicine, 2025, 40(1): 203-207.
[11] VERANI J R, MCGEE L, SCHRAG S J, et al. Preven-tion of perinatal group B Streptococcal disease-revised guidelines from CDC, 2010[J]. MMWR. Recommenda-tions and Reports, 2010, 59(RR/10): 1-36.
[12] 谢幸,孔北华,段涛.妇产科学[M].9版.北京:人民卫生出版社,2018:204-222. XIE X, KONG B H, DUAN T. Obstetrics and Gynecol-ogy[M]. 9th Ed. Beijing: People’s Medical Publishing House, 2018: 204-222.
[13] ROTTENSTREICH M, ROTEM R, BERGMAN M, et al. Assessment of maternal GBS colonization and early-onset neonatal disease rate for term deliveries: a decade perspective[J]. Journal of Perinatal Medicine, 2019, 47(5): 528-533.
[14] 刘娜,尤建萍,王博,等.妊娠晚期孕妇阴道B族链球菌感染青霉素钠治疗对血浆凝血功能及新生儿结局的影响分析[J].现代检验医学杂志, 2019, 34(4): 146-150. LIU N, YOU J P, WANG B, et al. Analysis on the influences of penicillin Sodium treatment on plasma coagulation functions and neonatal outcomes in late pregnant women with vaginal group B Streptococcus infections[J]. Journal of Modern Laboratory Medicine, 2019, 34(4): 146-150.
[15] WANG Z X, PU W Y, LIU Q, et al. Association of gut micro-biota composition in pregnant women colonized with group B Streptococcus with maternal blood routine and neonatal blood-Gas analysis[J]. Pathogens, 2022, 11(11): 1297.
[16] QIAO T C, TIAN H Y, SHAN S Z, et al. Causal rela-tionship between inflammatory factors and cerebral small vessel disease: univariate, multivariate, and sum-mary-data-based Mendelian randomization analysis[J]. Brain and Behavior, 2024, 14(2): e3399.
[17] SIMMONS D, IMMANUEL J, HAGUE W M, et al. Regression from early GDM to normal glucose toler-ance and adverse pregnancy outcomes in the treatment of booking gestational diabetes mellitus study[J]. Dia-betes Care, 2024, 47(12): 2079-2084.
[18] 徐军娟,倪娟,徐丽燕.孕晚期GBS感染不良妊娠结局影响因素及IL-8水平[J].中华医院感染学杂志, 2021, 31(2): 272-276. XU J J, NI J, XU L Y. Influencing factors for adverse pregnancy outcomes and IL-8 level of pregnant women with GBS infection in third trimester [J]. Chinese Jour-nal of Nosocomiology, 2021, 31(2): 272-276.
[19] 中华医学会围产医学分会.中华医学会妇产科学分会产科学组.预防围产期B族链球菌病(中国)专家共识[J].中华围产医学杂志,2021,24(8):561-566. Society of Perinatal Medicine,Chinese Medical Asso-ciation, Obstetrics Subgroup, Society of Obstetrics and Gynecology,Chinese Medical Association. Chinese experts consensus on prevention of perinatal group B Streptococcal disease [J]. Chinese Journal of Perinatal Medicine, 2021, 24(8): 561-566.
[20] LIU H, LI J, GUO J Y, et al. A prediction nomogram for neonatal acute respiratory distress syndrome in late-preterm infants and full-term infants: a retrospective study [J]. eClinical Medicine, 2022, 50: 101523.
[21] TANG J, GOU W L, FU Y Q, et al. Association be-tween postterm pregnancy and adverse growth out-comes in preschool-age children[J]. The American Journal of Clinical Nutrition, 2022, 116(2): 482-490.
[22] DENG K, LIANG J, MU Y, et al. Preterm births in Chi-na between 2012 and 2018: an observational study of more than 9 million women[J]. Lancet Global Health, 2021, 9(9): e1226-e1241.
[23] LIU Y X, LIU W L, ZHUANG G Y, et al. Colonisation of group B Streptococcus and its effects on pregnancy outcomes in pregnant women in Guangzhou, China: a ret-rospective study [J]. BMJ Open, 2023, 13(11): e078759.
[24] SAITO Y, KOBAYASHI S, IKEDA-ARAKI A, et al. Association between pre-pregnancy body mass index and gestational weight gain and perinatal outcomes in preg-nant women diagnosed with gestational diabetes mellitus:the Japan environment and children’s study[J]. Journal of Diabetes Investigation, 2022, 13(5): 889-899.
[25] ELORANTA A M, GUNNARSDOTTIR I, THORIS-DOTTIR B, et al. The combined effect of pre-pregnan-cy body mass index and gestational weight gain on the risk of pre-labour and intrapartum caesarean section-the ICE-MCH study[J]. PLoS One, 2023, 18(1): e0280060.
[26] FR?EN J F, ARNESTAD M, FREY K, et al. Risk factors for sudden intrauterine unexplained death: ep-idemiologic characteristics of singleton cases in Oslo, Norway, 1986-1995[J]. American Journal of Obstetrics and Gynecology, 2001, 184(4): 694-702.
[27] 梁嘉宁,孙悦,管梓妤.生育的集体动态:低生育率背景下生育意愿同伴效应研究[J].西华大学学报(哲学社会科学版), 2025, 44(2): 66-77. LIANG J N, SUN Y, GUAN Z Y. Collective dynamics of fertility: peer effects on fertility intentions in low-fer-tility contexts[J]. Journal of Xihua University (Philoso-phy & Social Sciences), 2025, 44(2): 66-77.
[28] SYSOEVA A P, NEPSHA O S, MAKAROVA N P, et al. Influence of extracellular vesicles from the follicular fluid of young women and women of advanced mater-nal age with different miRNA profiles on sperm func-tional properties[J]. Bulletin of Experimental Biology and Medicine, 2022, 173(4): 560-568.
[29] FANG H Y, LIU J, QIAN K, et al. Intestinal ultrasound combined with blood inflammatory markers is a more efficient tool in evaluating severity of crohn’s disease:a pilot study [J]. Gastroenterology Research and Prac-tice, 2023, 2023: 2173396.
[30] 冯宁,陈帅,窦明月,等.外周血IL-6、IL-10、SIRI在细菌性肺炎感染患者中的表达特征[J].临床医学研究与实践,2024,9(25):18-21. FENG N, CHEN S, DOU M Y, et al. Expression char-acteristics of IL-6, IL-10 and SIRI in peripheral blood of patients with bacterial pneumonia infection [J]. Clin-ical Research and Practice, 2024, 9(25): 18-21.
[31] OMAIR M, BAIG M S, FAROOQUI W A, et al. Re-lationship of neutrophil lymphocyte ratio, monocyte lymphocyte ratio and neutrophil monocyte ratio with treatment response in pulmonary tuberculosis patients during intensive phase treatment[J]. BMC Infectious Diseases, 2024, 24(1): 615.
[32] CURSI L, LANCELLA L, MARIANI F, et al. Mono-cyte-to-lymphocyte, neutrophil-to-lymphocyte and neu-trophil-to-monocyte plus lymphocyte ratios in children with active tuberculosis: a multicentre study[J]. Acta Paediatrica, 2023, 112(11): 2418-2425.
[33] 李革,余怡如,陈放,等.超敏CRP和单核细胞/淋巴细胞比值在妊娠糖尿病孕晚期发生不良妊娠结局中的价值研究[J].微循环学杂志, 2022, 32(4): 70-74. LI G, YU Y R, CHEN F, et al. Association of hypersen-sitive CRP and monocyte/lymphocyte ratio with adverse pregnancy out-comes in pregnant women with gesta-tional diabetes mellitus in the third trimester[J]. Chinese Journal of Microcirculation, 2022, 32(4): 70-74.
[34] AGAOGLU Z, TANACAN A, IPEK G, et al. Systemic inflammation response index and systemic immune-inflam-mation index for prediction of pregnancy loss: a case-control study[J]. Reproductive Sciences, 2024, 31(6): 1565-1572.

相似文献/References:

[1]郝晓燕,周 磊,白丛霞,等.基于机器学习方法建立血清多种肿瘤标志物联合区分胃炎和胃癌的三种模型及评价[J].现代检验医学杂志,2023,38(02):124.[doi:10.3969/j.issn.1671-7414.2023.02.023 ]
　HAO Xiao-yan,ZHOU Lei,BAI Cong-xia,et al.Establishment and Evaluation of Three Models for the Distinguish Gastritis and Gastric Cancer with Multiple Serum Tumor Markers Based on Machine Learning Methods[J].Journal of Modern Laboratory Medicine,2023,38(02):124.[doi:10.3969/j.issn.1671-7414.2023.02.023 ]
[2]曹俊杰,赵二川,王 瑞,等.基于GEO数据库采用生物信息学分析和机器学习筛选痛风中程序性细胞死亡相关关键基因及实验验证[J].现代检验医学杂志,2025,40(04):97.[doi:10.3969/j.issn.1671-7414.2025.04.017]
　CAO Junjie,ZHAO Erchuan,WANG Rui,et al.Identification and Experimental Validation of Programmed Cell Death- Related Key Genes in Gout Using Bioinformatics Analysis and Machine Learning Based on GEO Database[J].Journal of Modern Laboratory Medicine,2025,40(02):97.[doi:10.3969/j.issn.1671-7414.2025.04.017]
[3]饶亚妮,黄新周,卫永鲲.基于网络毒理学和生物信息学探索双酚A对骨性关节炎发生的潜在影响及关键基因筛选的分析研究[J].现代检验医学杂志,2026,41(02):160.[doi:10.3969/j.issn.1671-7414.2026.02.027]
　RAO Yani,HUANG Xinzhou,WEI Yongkun.Analytical Study of Bisphenol A Potential Impact on Osteoarthritis Development and Key Gene Screening Based on Network Toxicology and Bioinformatics[J].Journal of Modern Laboratory Medicine,2026,41(02):160.[doi:10.3969/j.issn.1671-7414.2026.02.027]

备注/Memo

备注/Memo:
基金项目：济宁市重点研发计划（No. 2023YXNS145）。
作者简介：于文凯（1989-），男，硕士，主管技师，研究方向：菌群与生殖健康方向的研究，E-mail：yuwenkai518@163.com。
通讯作者：郑康（1987-），女，本科，主管技师，研究方向：菌群与生殖健康方向的研究，E-mail：zhengkang0805@163.com。
更新日期/Last Update: 2026-03-15