Multi-agent system for early prediction of urinary bladder inflammation disease

This paper presents an efficient real-time knowledge base architecture for multi-agent based patient diagnostic system for chronic disease management, basically, the early detection of Inflammation of urinary bladder and Nephritis of renal pelvis origin diseases. The model integrates information stored heterogeneous and geographically distributed healthcare centers. The paper presents two main contributions. First, a proposed multi-agent based system for mining frequent itemsets in distributed databases. Second, the implementation of this model on distributed medical databases in order to generate hidden medical rules. The proposed model can gather information from each department or from different hospitals, and using the cooperative agents it analyzes the data using association rules as a data mining technique. The proposed model improves the diagnostic knowledge and discovers the diseases based on the minimum number of effective tests, thus, providing accurate medical decisions based on cost effective treatments. It can also predict the existence or the absence of the diseases, thus improving the medical service for the patients. The proposed multi-agent system constitute an effort toward the design of intelligent, flexible, and integrated large-scale distributed data mining system.