The thesis is devoted to the ways of reduction of perinatal morbidity and mortality in the high perinatal risk group by improving the prediction of complications, as well as optimization of non-invasive and invasive methods of fetal surveillance: echoscopy, dopplerography, cardiotocography, fetoscopy. Based on experimental studies, metabolic processes in acute and chronic fetal hypoxemia in an animal model (Rh?n sheep) were analyzed. A study of the quality and technical characteristics of ultrathin instruments used for invasion into the amniotic cavity (namely fiber-optic fetoscope with a diameter of 1 / 1.2 mm and puncture needles with a diameter of 26 / 29G) were hold. An optimized method of actography (taking into account the circadian rhythms of the fetus) and a semantic approach to the interpretation of CTG have been developed and implemented. The effectiveness of the methods has been proven in a clinical study. New criteria for assessing the fetal state were proposed, in particular: index of variability (ratio of episodes of low and high rhythm variability), cardiovascular index (ratio of short-term variability of heart rate and pulsation index in the venous duct of the fetus). Their effectiveness was proved in a clinical trial. For the first time, a multifactor model for predicting perinatal losses and angioplacental maternal complications has been developed on sufficient clinical material. Factors associated with the risk of perinatal loss include: the presence of chronic hypertension, preeclampsia in previous pregnancies, type of fertilization (natural or artificial), concentration of PAPP-A (MoM), concentration of free ?-HCG (MoM) in the second trimester of pregnancy, mean pulsation index in the uterine arteries at 28-30 weeks of pregnancy, pulsation index in the middle cerebral arteries at 28-30 weeks of pregnancy, episodes of low variability of fetal heart rate in the III trimester of pregnancy, episodes of high variability of fetal heart rate in the III trimester of pregnancy. The model, built on the selected features, allows with a sensitivity of 73.1% (95% CI 52.2% - 88.4%) and a specificity of 72.7% (95% CI 69.3% - 75.9%) to predict the risk perinatal loss. The accompanying factors that contribute to the occurrence of obstetric accidents have been identified and the degree of their impact has been calculated. For the first time, a comprehensive pregnancy monitoring program has been developed, which includes biochemical, cardiotocographic and echoscopic methods, and its effectiveness in reducing perinatal losses has been proven. Application of the developed algorithms, including the innovations proposed during the study, namely: multifactor prognostic model, subtle tools for invasive interventions, modified method of actography, semantic approach in cardiotocography, introduction of new combined criteria for fetal condition assessment (variability index, cardio-vascular index) to improve the diagnosis of late forms of growth restriction (OR 4.14 [1.42-12.09], p = 0.009), to reduce the frequency of urgent cesarean sections (OR 1.61 [1.03-2.49], p = 0.046) and to reduce perinatal mortality (OR 4.89 [1.09-21.3] , p = 0.041) due to the reduction of antenatal losses (OR 2.2 [1.06-4.337], p = 0.045). A new approach to the follow-up of patients with a high level of sFlt-1 / PlGF ratio has been proposed.
