Intelligent system for diabetes prediction in patients with chronic pancreatitis

Abstract

This paper presents the results of the JSM-method for automated support of scientific research (ASSR JSM-method) implemented in a computer intelligent system (IS-JSM), which predicts the development of diabetes in patients with chronic pancreatitis. For the first time the ASSR JSM-method is applied to a sequence of expanding databases of facts, which was used for the detection of empirical regularities (ERs), viz., preserved causes of the studied effect (development of diabetes in patients with chronic pancreatitis). To recognize ERs in the IS-JSM, we used an algebraic lattice of JSM-reasoning strategies (inductive inference rules). These results have an informative clinical interpretation and prove the usefulness of data mining using the IS-JSM, which can be used as a tool for evidence-based medicine.