Rule-based Medical Treatment Graph for the Modeling of Hypo- and Hyperglycemia at Onset

Abstract

This paper proposes a rule-based medical treatment graph (RB-MTG), a decision support tool that assists physicians in establishing insulin therapy. The RB-MTG models clinical pathways, i.e. the sequences of blood glucose measurements and insulin injections. It provides visualization of alternative clinical pathways, especially those that lead to dangerous states of the patient’s health. By interpreting the RB-MTG, the physician assesses the patient’s condition and plans their insulin therapy. At each phase of the treatment, the RB-MTG suggests the insulin dosage that leads to normoglycemia - the blood glucose level that is the norm for a healthy person. This way, it is possible to avoid the course of the disease that leads to hypo- or hyperglycemia. Physicians have verified the usefulness of our approach.