A New XGBoost Algorithm Based Prediction Model for Fetal Growth Restriction in Patients with Preeclampsia

Background: To construct a predictive model for fetal growth restriction (FGR) in preeclampsia (PE) patients using extreme Gradient Boosting (XGBoost) algorithm. Methods: A prospective study was conducted in the Obstetrics Department of Wuming Hospital from October 1, 2016, to October 1, 2021. A total of 303 preeclampsia patients were divided into two groups based on FGR status (restricted vs. unrestricted group). The clinical data and laboratory indicators between the two groups were compared. Logistics multivariate analysis and the XGBoost algorithm model were used to identify the risk factors for FGR in preeclampsia. Moreover, we used the receiver operating characteristic (ROC) curve to verify the accuracy of the XGBoost algorithm model. Results: Multivariate analysis and XGBoost algorithm modeling could predict the risk factors for FGR using clinical data and laboratory indicators. ROC analysis revealed that the area under the curve of the XGBoost algorithm model was 0.851, indicating a good fit. Conclusions: The XGBoost algorithm model can predict the occurrence of FGR in preeclampsia patients. The top three risk factors, triglyceride (TG), total cholesterol (TC), and lipoprotein (a) [Lp (a)], can be used as important predictors of poor patient prognosis in clinical settings.