Fuzzy Hybridization of "Artificial Neural Networks" (ANN) Based Signal and Image Processing Techniques: Application to Intelligent "Computer Aided Medical Diagnosis" (CAMD)

In this paper, an automated fault diagnosis system essentially based on neural networks and fuzzy logic, in a hybrid scheme, is suggested. First, a signal classification and image classification, resulting in a signal diagnosis and image diagnosis respectively, are developed. Such dual-classification is then exploited in a fuzzy system 1 to ensure a satisfactory reliability to medical diagnosis and particularly for auditory pathologies. Second, this reliability is reinforced using the obtained diagnosis result with an auditory threshold parameter of patients exploited in a fuzzy system 2 in order to generate the decision-making of the final diagnosis result. Finally a discussion is given with regard to the suggested hybrid approach and the decision-making phase for an automated fault diagnosis system.