Defining medical concepts by linguistic variables with fuzzy Arden Syntax

Diagnostic and therapeutic decisions in the domain of medicine usually depend on observed or measured data, such as observations made during an examination, or laboratory data. Furthermore, such decisions are often based on abstract domain knowledge. For decision-making, it is necessary to establish relationships between concrete data and abstract knowledge. These relationships can be defined in a formal way by fuzzy set theory. We extended the Arden Syntax for Medical Logic Systems, which can be used for decision support in the medical domain, to include concepts of fuzzy set theory. This paper presents extensions for defining linguistic variables by Arden Syntax, which can formalize different states of an abstract concept, such as different levels of blood glucose. Arden Syntax linguistic variables can be used within conditional expressions in decision rules or within fuzzy control rules for computer-aided diagnosis and therapy.