Artificial intelligence technologies in predicting preeclampsia

Abstract

The strategy for preserving reproductive potential in the Russian Federation is focused on the personalized women’s health care and based on preclinical identification of gynecological diseases and major obstetric syndromes at the stage of predicting adverse outcomes and subsequent preventive measures able to reduce maternal and perinatal morbidity and mortality, decrease women and neonatal disability as well as profoundly reduce extremely high costs on care of premature infants. The search for effective predictive methods of preeclampsia (PE) at the stage of preconception and in the first trimester of pregnancy is driven by the desire to identify women at greater risk of developing hypertensive disorders in order to take the necessary effective measures for preventing placental pathology as early as possible, thereby preventing or reducing incidence rate of PE. At the same time, identifying a group of high-risk women would allow to rationally plan prenatal care, timely recognize emergence of multiple organ dysfunction and immediately begin pathogenetic and symptomatic therapy. Taking into account the national and global experience of using predictive analytics software proving their success in reproductive medicine, it is reasonable to assume that converting prognosis into digital format by using artificial intelligence (AI) algorithms will open new opportunities for increasing accuracy of individual risk calculation for PE by meeting current paradigm of personalized preventive medicine. Our scientific review on domestic and international publications aims to inform a wide range of obstetricians-gynecologists about advances in AI technologies as well as prospects for machine learning to predict PE.