IT-technologies for detecting retinopathy of prematurity

Aim. Determination of effective clinical and anamnestic risk indicators for the development of retinopathy of prematurity for the development of an IT technology software product (calculator). Methods. A non-randomized prospective study of 507 premature infants with retinopathy of prematurity and a group at risk of developing retinopathy of prematurity (common risk factors for pathology of maternal obstetric history and somatic status of premature infants) was conducted in order to create a computer program for a calculator using the standard machine learning method of the Ridge model. Results. When analyzing the results of machine learning using the Ridge model program, the significance coefficients of risk factors for mother and child were identified. Based on the analysis of the short-term risk factors identified, 10 indicators from the obstetric history of the mother and the somatic status of the child were determined, which formed the basis for the development of a computer calculator for determining the risk of retinopathy of prematurity. Conclusions. The resulting program allows neonatologists and ophthalmologists to calculate the risk of developing retinopathy of prematurity in the first weeks of life and to carry out timely preventive measures. Key words: retinopathy of prematurity, calculator, computer program, IT technologies