Machine learning applied to birth type prediction

Abstract

Objective: To develop statistical models, using Machine Learning (ML), techniques, to model and predict care and administrative indicators in the UFC Hospital Complex. Methods: The study was carried out using the Monitoring/Evaluation Forms, where data is obtained from patients in labor who had normal delivery or cesarean section at the Obstetric Center. This data was used to generate the Machine Learning model. Results: The Decision Tree model achieved the highest accuracy compared to the other tested models, which indicates its ability to make correct predictions by a significant proportion. Conclusion: The model was successful in revealing that the type of cesarean delivery was the most prevalent mode of delivery among the study participants, corresponding to more than 50% of all deliveries. The possibility of extending the project to other universities would have a positive impact, as it could provide valuable information to improve decision-making regarding the type of delivery in different health contexts.