胎儿生长受限评估的混合机器学习模型:一项5年的随访研究

Hybrid machine learning models for fetal growth restriction assessment: a 5-year follow-up study

摘要

目的 探究列线图、决策树及随机森林3种混合机器学习模型对胎儿生长受限(FGR)的预测价值。方法 回顾性选取2018年1月至2022年12月南宁市红十字会医院行定期产检并分娩合并FGR孕妇111例,另选取本院同期未合并FGR孕妇87例.收集所有孕妇的临床资料、彩色多普勒超声指标、实验室检查指标、胎儿情况.采用多因素Logistic回归分析孕妇合并FGR的影响因素;以独立影响因素作为预测变量,构建列线图、决策树及随机森林3种预测模型,采用准确度、灵敏度、特异度、精确度、召回率、F1值及受试者工作特征(ROC)曲线评价模型的预测性能。结果 多因素Logistic回归分析结果显示:未规律补充叶酸、妊娠期高血压疾病(HDP)、大脑中动脉阻力指数(RI)(截断值0．86)、低空腹静脉血管内皮生长因子(VEGF)(截断值49．85ng/mL)及低脐带静脉血VEGF(截断值67．10ng/mL)是孕妇FGR发生的独立影响因素(OR值介于0．013~57．563之间,P＜0．05).基于独立危险因素分别构建三种模型的曲线下面积(AUC)、灵敏度、特异度,列线图模型为0．903(0．823~0．982)、0．925、0．750;决策树模型为0．894(0．809~0．979)、0．875、0．850;随机森林模型为0．968(0．931~1．000)、0．875、0．950.随机森林模型预测的AUC值高于决策树模型(Z＝-2．582,P＜0．05)。结论 列线图、决策树及随机森林预测模型对FGR的预测结果均具有较高的区分度,其中随机森林模型预测孕妇并发FGR的效果优于决策树模型,随机森林模型与列线图模型的预测效能无显著差异,两者可相互补充应用.

Abstract

Objective To explore the predictive value of three hybrid machine learning models (nomogram, decision tree and random forest) for fetal growth restriction (FGR). Methods A total of 111 pregnant women with FGR who underwent regular obstetric examination and delivered in Nanning Red Cross Hospital from January 2018 to December 2022 were retrospectively selected, and another 87 pregnant women without FGR in the same period of time were also selected. Clinical data, color Doppler ultrasound indexes, laboratory test indexes, and fetal conditions of all pregnant women were collected. Multifactorial Logistic regression was used to analysis the influencing factors of FGR in pregnant women. Three prediction models, nomogram, decision tree and random forest, were constructed using independent influencing factors as predictor variables. The predictive performance of the models was evaluated using accuracy, sensitivity, specificity, precision, recall, F1 value, and receiver operating characteristic (ROC) curves. Results Multifactorial logistic regression analysis showed that irregular folic acid supplementation, hypertensive disorders of pregnancy (HDP), middle cerebral artery RI (cut-off value=0.86), low fasting venous blood VEGF (cut-off value=49.85 ng/mL), and low umbilical cord venous blood VEGF (cut-off value=67.10 ng/mL) were independent factors for the occurrence of FGR in pregnant women (OR=0.013-57.563, respectively, all P<0.05). Based on the independent risk factors, the area under the curve (AUC), sensitivity, and specificity of the three models were constructed respectively, and were 0.903 (0.823-0.982), 0.925, 0.750 for the nomogram model; were 0.894 (0.809-0.979), 0.875, 0.850 for the decision tree model; were 0.968 (0.931-1.000), 0.875, 0.950 for the random forest model. The AUC value of random forest model was higher than that of decision tree model (Z=-2.582, P<0.05). Conclusion The prediction results of nomogram, decision tree and random forest prediction models on FGR all have a high degree of differentiation, among which the random forest model is better than the decision tree model in predicting. There is no significant difference in the prediction efficacy between the random forest model and the nomogram model, and the two can be complementary to each other in the application of predicting FGR.

关键词

胎儿生长受限 / 影响因素 / 列线图 / 决策树 / 随机森林 / 预测

Key words

fetal growth restriction / influence factor / nomogram / decision tree / random forest / prediction

参考文献

[1] Melamed N, Baschat A, Yinon Y, et al. FIGO (international federation of gynecology and obstetrics) initiative on fetal growth: best practice advice for screening, diagnosis, and management of fetal growth restriction[J]. Int J Gynaecol Obstet, 2021, 152 Suppl 1(Suppl 1): 3-57.

[2] Baschat AA, Galan HL, Lee W, et al. The role of the fetal biophysical profile in the management of fetal growth restriction[J]. Am J Obstet Gynecol, 2022, 226(4): 475-486.

[3] Figueras F, Caradeux J, Crispi F, et al. Diagnosis and surveillance of late-onset fetal growth restriction[J]. Am J Obstet Gynecol, 2018, 218(2S): S790-S802.e1.

[4] Sotiriadis A, Figueras F, Eleftheriades M, et al. First-trimester and combined first-and second-trimester prediction of small-for-gestational age and late fetal growth restriction[J]. Ultrasound Obstet Gynecol, 2019, 53(1): 55-61.

[5] 李玲, 张盼盼, 张文琴. 基于彩色多普勒超声定量参数列线图模型预测胎儿生长受限的临床价值[J]. 临床超声医学杂志, 2023, 25(2): 90-94.

[6] Ferreira-Santos D, Amorim P, Silva Martins T, et al. Enabling early obstructive sleep apnea diagnosis with machine learning: systematic review[J]. J Med Internet Res, 2022, 24(9): e39452.

[7] 谢幸, 孔北华, 段涛, 等. 妇产科学[M].9版. 北京: 人民卫生出版社, 2018.

[8] 段然, 漆洪波. 胎儿生长受限的规范化诊断和管理[J]. 实用妇产科杂志, 2021, 37(11): 801-803.

[9] 高文凯, 王瑞玲, 王凤梅, 等. 孕晚期宫内生长受限胎儿彩色多普勒血流显像相关参数变化及临床意义[J]. 湖南师范大学学报(医学版), 2021, 18(1): 150-154.

[10] Crispi F, Miranda J, Gratacós E. Long-term cardiovascular consequences of fetal growth restriction: biology, clinical implications, and opportunities for prevention of adult disease[J]. Am J Obstet Gynecol, 2018, 218(2S): S869-S879.

[11] 刘红青. 河南省封丘县590名育龄期女性叶酸增补相关知信行水平调查分析及干预策略[J]. 中国药物与临床, 2021, 21(19): 3265-3267.

[12] Lee S, Gleeson JG. Closing in on mechanisms of open neural tube defects[J]. Trends Neurosci, 2020, 43(7): 519-532.

[13] 陈甘讷, 蒋敏, 黄伟雯, 等. 围孕期叶酸补充情况及其对妊娠结局影响[J]. 中华疾病控制杂志, 2021, 25(2): 160-164.

[14] Alvestad S, Husebye ES N, Christensen J, et al. Folic acid and risk of preterm birth, preeclampsia, and fetal growth restriction among women with epilepsy: a prospective cohort study[J]. Neurology, 2022, 99(6): e605-e615.

[15] 姜虹羽, 庄秀娟, 蔡思铭, 等. 胎儿生长受限与脐动脉血流异常的相关性及影响胎儿生长的相关影响因素分析[J]. 河北医学, 2022, 28(8): 1379-1383.

[16] 彭翠, 贺惠琴, 佘芹, 等. 子痫前期并发胎儿生长受限的妊娠结局和危险因素分析[J]. 中国计划生育和妇产科, 2022, 14(1): 70-75.

[17] 洪海燕, 刘宇杰, 方丽妮, 等. 超声检查子宫动脉、脐动脉血流参数对胎儿生长受限的诊断价值及临床监测意义[J]. 陕西医学杂志, 2023, 52(5): 567-570.

[18] 张中华, 迟玉君, 李辉, 等. 孕晚期宫内生长受限胎儿与宫内正常发育胎儿超声多血管血流参数的差异性分析[J]. 中国医学装备, 2023, 20(1): 87-92.

[19] 杨俊娟, 郭华峰, 柳蕊. VitB12、叶酸、同型半胱氨酸和血管内皮生长因子水平与胎儿生长受限的相关性[J]. 中国妇幼健康研究, 2023, 34(6): 12-16.

[20] 朱念念, 赵敏, 夏闪. 妊娠期高血压疾病患者孕晚期血清PLGF、VEGF水平与胎儿生长受限的关系[J]. 山东医药, 2022, 62(31): 31-34.

[21] Deval R, Saxena P, Pradhan D, et al. A machine learning-based intrauterine growth restriction (IUGR) prediction model for newborns[J]. Indian J Pediatr, 2022, 89(11): 1140-1143.

基金资助

广西壮族自治区卫生健康委员会自筹经费科研课题(Z20201197)