Title :
Image Recognition Based Analysis and Comparison of Hybrid Intelligent Approaches for Computer Aided Diagnosis
Author :
Chohra, Amine ; Kanaoui, Nadia ; Madani, Kurosh
Author_Institution :
Paris XII Univ., Lieusaint
Abstract :
In this paper, two hybrid intelligent approaches are suggested for computer aided diagnosis systems in a biomedicine application: auditory diagnosis based on auditory brainstem response test. Indeed, these approaches are developed through the hybrid intelligent system 1 (HIS_1) and the hybrid intelligent system (HIS_2), based on neural classification and fuzzy decision-making of global image and subdivided image. In fact, in each system, a double classification is exploited in a primary diagnosis fuzzy system to ensure a certain degree of reliability. This reliability is reinforced using a confidence parameter with the primary diagnosis results (from primary diagnosis fuzzy system), exploited in a final diagnosis fuzzy system, in order to generate the appropriate diagnosis with a confidence index (CI). Afterwards, experimental set up and auditory diagnosis results are presented for HIS_1 and HIS_2, analyzed and compared based on image recognition.
Keywords :
fuzzy neural nets; image classification; image recognition; knowledge based systems; medical image processing; auditory brainstem response test; auditory diagnosis; biomedicine; computer aided diagnosis systems; confidence index; fuzzy decision-making; global image; hybrid intelligent system; image recognition; neural classification; primary diagnosis fuzzy system; subdivided image; Application software; Biomedical computing; Decision making; Fuzzy systems; Hybrid intelligent systems; Image analysis; Image recognition; Knowledge representation; Pattern recognition; Signal analysis; Decision Support; Fuzzy Decision-Making; Image Knowledge Representation; Learning And Adaptation; Neural Classification;
Conference_Titel :
Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications, 2007. IDAACS 2007. 4th IEEE Workshop on
Conference_Location :
Dortmund
Print_ISBN :
978-1-4244-1347-8
Electronic_ISBN :
978-1-4244-1348-5
DOI :
10.1109/IDAACS.2007.4488441