Type-2 Fuzzy Relations: An Approach towards Representing Uncertainty in Associative Medical Relationships

The acquisition of precise values such as symptoms, signs, laboratory test results, and diseases/diagnoses for expressing meaningful associative relationships between medical entities has always been regarded as a critical part of developing medical knowledge-based systems. After the introduction of fuzzy sets, researchers became aware of the fact that a central problem in the use of fuzzy sets is constructing the membership function values. The complication arises from the uncertainty associated with assigning an exact membership grade for each element within the considered fuzzy set. Type-2 fuzzy set handles this problem by allocating a different fuzzy set to each element. This paper addresses the subject of medical knowledge acquisition and representation by proposing consistent interval type-2 fuzzy relations in the context of fuzzy inclusion as a measure of representing the degrees of association between medical entities. The concept of interval type-2 fuzzy relation will be introduced to represent the uncertainty and vagueness between medical entities.