Multi-Agent based Diagnostic Model for Diabetes

With the advent of Agent Technology, and the challenges facing the health care sector and also the distributive and complex nature of the activities in the medical domain, there is need to use the new programming paradigm in order is to explore, develop, design and implement a prototype system for multi-agent diabetes diagnosis that finds the possibility of illness based on the patient’s symptoms which will aid in solving the problems as a result of wrong diagnoses especially in the developing countries like Nigeria. Fuzzy Rule Based Reasoning was used to build the knowledge based of the system based on the studied data of the diabetic patient. The prototype system is capable of diagnosing (Predicting) diabetes without measuring the glucose level of the patient and monitoring diabetic patient by providing them with the required amount of insulin based on the patient weight to enable the patients send an update to the doctors. The experimental result of the diagnosing aspect of the system shows that the accuracy of the system is 0.75 while sensitivity and specificity of the system is 0.88 and 0.5 respectively. Keyword: Fuzzy, Polyurea, Polydipsia, Polyphagia, Agent ——————————  ——————————