Abstract
The capacity to forecast nonreassuring fetal heart (NFH) is essential for minimizing perinatal complications; therefore, this research aims to establish if a machine learning (ML) model can predict NFH. This was a retrospective analysis of information gathered from singleton cases over the gestational age of 28 weeks that sought vaginal delivery between January 2020 and January 2022. The information was acquired from the "Iranian Maternal and Neonatal Network."A predictive model was built using four statistical ML models (decision tree classification, random forest classification, extreme gradient boost classification, and permutation feature classification with k-nearest neighbors). Because of the limited studies on the identification of NFH predictors, we decided to use the Chi-Square test to compare demographic, obstetric, maternal, and neonatal factors to identify NFH predictors. Then, all variables with p-values less than 0.05 were considered potential NFH predictors. The area under the receiver operating characteristic curve (AUROC), accuracy, precision, recall, and F1 score were measured to evaluate their diagnostic performance. The incidence of NFH in our study population was 9.2%. Based on our findings NFH was more common in cases of intrauterine growth restriction, late-term, post-term, and preterm births, preeclampsia, placenta abruption, primiparous, induced labor, male fetus, and lower in birth with the presence of doula support. Random forest classification (AUROC: 0.77), decision tree classification and extreme gradient boost classification (AUROC: 0.76), and permutation feature classification with K-nearest neighbors (AUROC: 0.77), all showed good performance in predicting NFH. The higher performance belonged to random forest classification with an accuracy of 0.77 and precision of 0.72. Although this study found that the classification tree models performed well in predicting NFH, more research is needed to make a better conclusion on the performance of ML models in predicting NFH.