USE OF NEURAL NETWORKS FOR PREDICTION OF COMPLICATIONS OF THE GESTATION PERIOD

The article discusses in detail the development, use and use of neural networks for practical healthcare, especially in predicting the development of complications and preventing the development of complications. This is especially relevant now given the expected risk of adverse outcomes during pregnancy. Neural networks are obtained by themselves, simple elements are built among themselves - formal neurons. Most of the work on neuroinformatics is devoted to the transfer of various algorithms for solving problems in such networks. The core of the ideas used is the idea that neurons can be modeled by fairly simple automata, and the complexity of the brain, the flexibility of its functioning, and other important qualities are determined by the connections between neurons. The purpose of this study: was the development of neural network expert programs as the most important method for researching medical data. Our study was the result of testing the possibility of using neural networks in obstetrics and gynecology to predict gestational complications and the development of placental insufficiency. The definition of the risk group was carried out on the basis of the model developed by us. A neural network for prediction was developed, the possibility of using this model in obstetrics and gynecology was tested to predict gestational complications in 1024 women with a high risk of developing this pathology. Confirming the effectiveness of the proposed model, based on the data obtained from a prospective study, we calculated the indicators of its sensitivity, specificity, positive and negative predictive value. The result of the study confirms that the trained neural network makes a reliable prediction of the development of gestational complications and the development of insufficiency of the placental complex. As a result of the research, an important conclusion can be drawn that neural networks are the most important method for studying medical data, universal and highly effective, focused on active use in practical medicine.