Zobraziť v EDS

Interpretable machine learning model for predicting low birth weight in singleton pregnancies: a retrospective cohort study

Uložené v:
Podrobná bibliografia
Názov: Interpretable machine learning model for predicting low birth weight in singleton pregnancies: a retrospective cohort study
Autori: Xiaojuan Wu, Qingxiang Zhao, Yong Gao, Yiyu Zhang, Linrui Xu, Xianzhu Cong, Na Sun, Fuyan Shi, Suzhen Wang
Zdroj: BMC Pregnancy and Childbirth, Vol 25, Iss 1, Pp 1-16 (2025)
Informácie o vydavateľovi: BMC, 2025.
Rok vydania: 2025
Zbierka: LCC:Gynecology and obstetrics
Predmety: Low birth weight, Machine learning, Shapley additive explanations, Nomogram, Neonatal complications, Gynecology and obstetrics, RG1-991
Popis: Abstract Background Low birth weight (LBW), defined as a newborn weighing less than 2500 g, is an increasingly significant public health concern. Exploring the risk and protective factors for LBW is getting more and more important. This study aimed to utilize predictive models to identify the critical factors associated with LBW in singleton pregnancies. Methods A retrospective cohort study was conducted at the Binzhou Medical University Hospital, China, from 2022 to 2023. Singleton pregnancies with gestational age exceeding 27 weeks were included, while multiple pregnancies and fetal anomalies were excluded. Logistic regression (LR) model and four machine learning (ML) algorithms were tested(random forest, light gradient boosting machine, support vector machine, and extreme gradient boosting). The LR model was interpreted through odds ratio analysis and clinical nomogram visualization. Shapley Additive Explanations (SHAP) analysis was used to interpret the importance and impact of individual features on ML model. Results In this cohort of 10,227 deliveries, 237 cases were classified as LBW. The XGBoost model exhibited superior performance in predicting LBW, with an AUC of 0.786 (training set) and 0.741 (test set, ) AUPRC of 0.350 and 0.186, respectively. Both LR and XGBoost model identified maternal age, gestational age, BMI, hypertensive disorders of pregnancy (HDP), fetal distress as the critical factor. Additionally, a follow-up study on LBW found that LBW infants prone to significant health challenges, such as a high rate of hospitalization and the complex conditions including congenital anomalies, neonatal respiratory distress syndrome (NRDS) and neonatal hyperbilirubinemia. Conclusion This study demonstrated that the LR and XGBoost model exhibited clinically meaningful predictive performance in identifying factors associated with LBW in singleton pregnancies. Pregnant women with a gestational age of less than 37 weeks, a gestational BMI below 18 kg/m², maternal age under than 25 years, and maternal comorbidities such as HDP or fetal distress are at higher risk of delivering LBW infants.
Druh dokumentu: article
Popis súboru: electronic resource
Jazyk: English
ISSN: 1471-2393
Relation: https://doaj.org/toc/1471-2393
DOI: 10.1186/s12884-025-08318-0
Prístupová URL adresa: https://doaj.org/article/3e2cea1dea4f4458b573dd4901029be1
Prístupové číslo: edsdoj.3e2cea1dea4f4458b573dd4901029be1
Databáza: Directory of Open Access Journals
Popis
Abstrakt:Abstract Background Low birth weight (LBW), defined as a newborn weighing less than 2500 g, is an increasingly significant public health concern. Exploring the risk and protective factors for LBW is getting more and more important. This study aimed to utilize predictive models to identify the critical factors associated with LBW in singleton pregnancies. Methods A retrospective cohort study was conducted at the Binzhou Medical University Hospital, China, from 2022 to 2023. Singleton pregnancies with gestational age exceeding 27 weeks were included, while multiple pregnancies and fetal anomalies were excluded. Logistic regression (LR) model and four machine learning (ML) algorithms were tested(random forest, light gradient boosting machine, support vector machine, and extreme gradient boosting). The LR model was interpreted through odds ratio analysis and clinical nomogram visualization. Shapley Additive Explanations (SHAP) analysis was used to interpret the importance and impact of individual features on ML model. Results In this cohort of 10,227 deliveries, 237 cases were classified as LBW. The XGBoost model exhibited superior performance in predicting LBW, with an AUC of 0.786 (training set) and 0.741 (test set, ) AUPRC of 0.350 and 0.186, respectively. Both LR and XGBoost model identified maternal age, gestational age, BMI, hypertensive disorders of pregnancy (HDP), fetal distress as the critical factor. Additionally, a follow-up study on LBW found that LBW infants prone to significant health challenges, such as a high rate of hospitalization and the complex conditions including congenital anomalies, neonatal respiratory distress syndrome (NRDS) and neonatal hyperbilirubinemia. Conclusion This study demonstrated that the LR and XGBoost model exhibited clinically meaningful predictive performance in identifying factors associated with LBW in singleton pregnancies. Pregnant women with a gestational age of less than 37 weeks, a gestational BMI below 18 kg/m², maternal age under than 25 years, and maternal comorbidities such as HDP or fetal distress are at higher risk of delivering LBW infants.
ISSN:14712393
DOI:10.1186/s12884-025-08318-0