Возможности ранней диагностики гипоксического повреждения головного мозга у доношенных новорожденных

Abstract

The aim of the research. To create a technology for early diagnosis of hypoxic brain damage in full-term newborns. Material and methods. A retrospective analysis of birth histories and cytokine levels in 89 newborns in 2017-2020 was carried out at the premises of perinatal centres in Chita. A total of two study groups were outlined: group 1 included 39 newborns with chronic hypoxia; group 2 included 50 newborns with birth asphyxia. The levels of interleukins (IL-1β, IL-4, IL-6, IL-8, IL-10), TNF-α and NSE in the of umbilical cord blood serum were measured via the immunoenzymatic method. Statistical processing of the results was carried out using the IBM SPSS Statistics Version 25.0 soft ware package (International Business Machines Corporation, USA). Results. The structure of the trained neural network included 6 input neurons refl ecting the level of relevant interleukins (IL-1β, IL-4, IL-6 and IL-8) and the Apgar score at the end of the first and fifth minutes. The percentage of incorrect predictions of the obtained neural network was 17.4 %. Conclusion. A complex approach based on integration of the Apgar score at the end of the first and fifth minutes of life and the level of the cytokines IL-1β, IL-4, IL-6 and IL-8 in umbilical cord blood in the structure of the neural network allows the diagnosis of hypoxic brain damage development with an accuracy of 82.6 %. Application of this technology in clinical practice will make it possible to timely diagnose the pathology of the central nervous system and reduce the frequency of adverse neurological outcomes.