FNDSB: A fuzzy-neuro decision support system for back pain diagnosis

Abstract Back pain is a common pain felt in the back. In the world, it is the fifth common reason for physician visits and in the U.S. about 90% of human adults have back pain at some time in their life. In this paper, we proposed a decision support system for intelligent diagnosis of back pain using a fuzzy-neuro technique to manage the fuzzy concepts. It is also provide an intelligent decision support platform that can assist physicians to diagnose and produce accurate medical advices. The proposed system consists of a user interface that receive symptoms and produce accurate diagnosis with its severity, a fuzzy inference system which contains the strategies of reasoning process based on fuzzification and defuzzification techniques, a fuzzy-neuro system that composed of neural network with fuzzy logic concepts, and a knowledge base which collects of linguistic fuzzy rules. The FNDSB describe knowledge acquisition and representation methods, a way of production linguistic fuzzy rules organization, fuzzification and defuzzification of clinical parameters as input and output values using a Triangular Membership Function (TMF) and Centroid of Area (CoA) techniques respectively. The FNDSB was evaluated using a case study of 10 patients from Al-diwaniyah teaching hospital. According to the evaluation results the system performance around 83.6% efficiency in producing accurate back pain diagnosis.