The possibilities of neural network data analysis in predicting preterm labour

Preterm labour is one of the main causes of perinatal mortality and morbidity. To date, there are no effective ways to prevent preterm labor, which determines the relevance of this study. Therefore, the effectiveness of assessing the likelihood of their development determines the timeliness of therapy, which largely determines the outcome of pregnancy as a whole. A retrospective analysis of 220 birth histories for 2020 year was carried out on the basis of the perinatal center of the Regional Clinical Hospital. The total sample was divided into 2 study groups: group 1 icluded 20 patients who had spontaneous premature birth; Group 2 included 200 patients the birth of which occurred at full-term pregnancy. Statistically significant parameters such as: living in a rural or urban area, the presence of anemia, diseases of the urinary system, polyhydramnios, erythrocyte sedimentation rate, blood glucose, myelocytes and international normalized ratio on the eve of childbirth were included in the test database, which formed the basis training a multilayer perceptron. The structure of the trained neural network included 12 input neurons, one hidden layer containing 8 units, and 2 output neurons (Se=0,98, Sp=0,98, AUC=0,99 [95% CI 0,97-1,00], p