ПРИМЕНЕНИЕ ИСКУССТВЕННОГО ИНТЕЛЛЕКТА ДЛЯ ПРОГНОЗИРОВАНИЯ ПРЕЭКЛАМПСИИ И ЗАДЕРЖКИ РОСТА ПЛОДА: ИНТЕГРАЦИЯ КЛИНИЧЕСКИХ, УЛЬТРАЗВУКОВЫХ И БИОМОЛЕКУЛЯРНЫХ ДАННЫХ

Abstract

Preeclampsia and fetal growth restriction (FGR) remain leading causes of maternal and perinatal morbidity and mortality worldwide. The limited predictive value of traditional methods is attributed to the complex and multifactorial pathogenesis of these conditions, including impaired placentation, endothelial dysfunction, immune dysregulation, and angiogenic imbalance. In recent years, artificial intelligence (AI) has emerged as a powerful tool in obstetrics, enabling the analysis of large datasets and identification of hidden patterns. The aim of this study is to evaluate the application of machine learning techniques for predicting preeclampsia and FGR through the integration of clinical, ultrasound, biochemical, and immunological data. Modern algorithms, including neural networks, gradient boosting, and random forest models, are analyzed in terms of their diagnostic performance. AI-based models can improve prediction accuracy up to 85–92%, significantly outperforming conventional approaches. The integration of multi-omics data supports the development of personalized medicine in obstetrics and offers new opportunities for early diagnosis and prevention of pregnancy complications.