Forecasting Risk Pregnancies in Peru using Machine Learning

In Peru, between 2017 and 2021, there has been an upward trend in maternal deaths attributed to preeclampsia, reaching an average of 393 cases. Recognizing the importance of early detection of risks associated with preeclampsia, a predictive model was developed using machine learning techniques to forecast pregnancies at risk of preeclampsia. Six models (Support Vector Machine, C5.0 Algorithm, Classification and Regression Trees, k-Nearest Neighbors, Naive Bayes, Logistic Regression) were employed on 44,633 records analyzed from an obstetrics section located in a hospital in Peru, with 11 out of 24 variables selected, including laboratory tests. Logistic Regression demonstrated the best performance, achieving an AUC of 97.6% and an accuracy of 81.8%. This model has the potential to be implemented in hospital settings to detect preeclampsia early, which could significantly contribute to reducing maternal and neonatal mortality rates in the country.