ARTIFICIAL INTELLIGENCE IN THE PREDICTION OF PREECLAMPSIA: AN EMERGING TOOL FOR SCREENING AND EARLY DIAGNOSIS

Abstract

Preeclampsia is one of the leading causes of maternal and fetal morbidity and mortality, making early diagnosis essential for effective prophylactic interventions. Traditional screening methods often present limitations in sensitivity, which may result in delayed diagnosis of the condition. Therefore, this study aims to explore the impact of Artificial Intelligence (AI) and Machine Learning (ML) as emerging tools in the prediction of preeclampsia. The methodology was based on a narrative literature review, involving the selection and analysis of recent studies addressing the integration of clinical, laboratory, genomic, and hemodynamic variables. The main findings indicate that AI models achieve accuracy rates above 85%, outperforming conventional methods, which rarely exceed 60%. Algorithms such as Random Forest, Neural Networks, and Classification Trees have shown effectiveness in processing data such as blood pressure (systolic, diastolic, and mean), history of hypertension, angiogenic biomarkers (PlGF and sFlt-1), and noninvasive hemodynamic data in the first trimester. Institutional reports suggest a potential reduction in severe complications and improved risk stratification with the integration of AI systems into electronic health records. In conclusion, AI offers significant contributions by enabling continuous monitoring, personalized care, and early indication of prophylactic treatments, redefining paradigms in contemporary obstetric practice.