Supporting System for the Diagnosis of Diabetes Mellitus Based on Glucose Tolerance Test Responses Using a Fuzzy Inference

Diabetes Mellitus is one of the most common diseases in adults. This disease is characterized by an elevated blood glucose (BG) level due to absolute or relative insulin deficiency. Diagnosis of this disease is based on the responses to the 75g oral glucose tolerance test. In the criteria of the WHO, if the value of BG at two hours after glucose load (BG2h) is more than 200 mg/dl or if the value of fasting blood glucose (FBG) is more than 140mg/dl, then the patients is diagnosed as “Diabetes Mellitus (DM)”. If FBG is less than 139 mg/dl, the BG value at one hour is lower than 160 mg/dl and BG2h is less than 139mg/dl, the patient is normal (NOR). The boundary region between DM and NOR is known as “Impaired Glucose Tolerance (IGT)”. In this criteria a patient, for example, is diagnosed as DM or non DM. But the real data is a fuzzy set in every aspect. In this paper, responses of glucose Tolerance Test were examined for the diagnosis of diabetes mellitus. Our study proposed the concept of diabetes mellitus as fuzzy set and a new diagnostic system for diabetes using a fuzzy inference with the time course.