Neuro-fuzzy model and precedent expert system of the transformational hybrid intelligence for predicting outcomes of the acute pancreatitis

S. B. Rumovskaya

Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation

Abstract: In polymorbid and polyetiological diseases with complex clinical presentations and phasic progression such as acute pancreatitis, the development of infectious complications combined with inadequate treatment strategies and surgical interventions can lead to mortality rates reaching 100%. Improving outcomes for patients with acute pancreatitis requires the early prediction of infectious complications and potential lethal outcomes to identify the most severe patient cohort for timely comprehensive intensive therapy and adequate surgical treatment. Current diagnostic methods and approaches are maximally informative only by the end of the second to third days. However, the highest therapeutic efficacy is observed within the first 24–48 h. Thus, transitioning to personalized integrated prediction algorithms is critical. The paper presents the results of the developing a heterogeneous field of models within the transformational model designed as a "neuro-fuzzy – case-based expert system. This model is the part of the future cooperative self-adjusting hybrid intelligent systems for personalized patient state assessment (the case of acute pancreatitis).

Keywords: hybrid intelligent system, neuro-fuzzy system, precedent expert system, predicting in medicine, acute pancreatitis.

Received: 03.07.2025
Accepted: 15.10.2025

DOI: 10.14357/08696527250403