Artificial Intelligence in Predicting Cesarean Delivery: A Pilot Study

Abstract

Background: Cesarean section is among the most prevalent surgeries globally and its increasing prevalence has become a significant health issue of concern to people. Earlier detection of women who are at a higher risk of cesarean delivery can be achieved and lead to better intrapartum care and maternal and neonatal outcomes. The possibilities of artificial intelligence provide new methods of creating models of data predictions after analyzing the commonly available clinical variables.

Objective: To develop and evaluate a preliminary artificial intelligence–based model for predicting cesarean delivery in a pilot cohort of women presenting in labor.

Methodology: This pilot study is a prospective trial observational pilot study that took place at Multicenter between January 2022 and January 2023. A total of 72 women with singleton term pregnancies who were in active labor were enrolled. A structured proforma was used to measure the demographic, obstetric, and intrapartum variables. A logistic regression model, which is supported by AI was designed to predict mode of delivery. The measures used to determine the model performance were accuracy, sensitivity, specificity, and the area under the receiver operating characteristic curve.

Results: Of the 72 participants, 30 (41.7%) underwent cesarean delivery and 42 (58.3%) delivered vaginally. Previous cesarean section, nulliparity, and lower Bishop score were independently associated with increased odds of cesarean delivery. The prediction model achieved an accuracy of 78.0% and an area under the ROC curve of 0.84, indicating good discriminatory performance.

Conclusion: This pilot study demonstrates the feasibility of developing an artificial intelligence–based model to predict cesarean delivery using routine clinical data. The encouraging preliminary performance supports further large-scale, multicenter studies to validate and refine AI-assisted decision support tools in obstetric practice.

Keywords: Artificial intelligence; Cesarean delivery; Prediction model; Pilot study; Obstetrics.