Early prediction of sepsis using a proprietary application developed based on machine learning (artificial intelligence) : Abstract of the Ph.D. thesis in medical sciences: 321.19 – Anaesthesiology and intensive care

Abstract

The topicality and importance of the problem addressed. Two important aspects can be highlighted that outlines the research direction reflected in the paper: (a) The problem of sepsis as a variety of critical conditions, often difficult to diagnose in time, and the results of treatment depend closely on the time when treatment is started (with antibiotics) and the direct influence of these factors on mortality [1], which over time has decreased insignificantly [2] and (b) The emergence of a new player - the so-called artificial intelligence (AI) technologies, which often claim to be panaceas for many problems, including medical ones. The current research attempts to assess the possibility of using these technologies to address important issues in the management of this group of patients through early prediction (a few hours before onset) of sepsis, which would mitigate the delay in starting the treatment. [...]