Document Type : Original Research (Full Papers)
Department of Electrical engineering,Biomedical engineering and computer, Faculty of Engineering , Qazvin Branch,Islamic Azad University, Qazvin ,Iran
Department of Electrical engineering, Biomedical engineering and computer, Faculty of Computer and Information Technology Engineering, Qazvin Branch, Islamic Azad University, Qazvin, Iran
In recent years, the use of MRI images has been very much considered due to their high clarity and high quality in the diagnosis and determination of brain tumor and its features. In this study, to improve the performance of tumor detection, we investigated comparative approach of the different classifiers to select the most appropriate classifier for identifying and extracting abnormal tissue and selected the best one by comparing their detection accuracies rate. In this research, GLCM and GLRM methods are used to extracting discriminating features. Thus results in they reduce the computational complexity. fuzzy entropy measurement method is used to determine the optimal properties and finally, we compared the four FFNN, MLP, BPNN, ANFIS neural networks to perform the decision making and classification process. The purpose of these four neural networks are to develop tools for discriminating the malignant tumors from benign ones assisting deciding in clinical diagnosis. Based on the results, we achieved high results among all classifiers. The proposed methodology results in accurate and speedy detection of tumor in brain along with identiﬁcation of precise location of the tumor. In our opinion, the use of these classifiers can be very useful in the diagnosis of brain tumors in MRI images. Our other goal is to prove the suitability of the ANN method as a valuable method for statistical methods. The novelty of the paper lies in the implementation of the proposed method for discriminating the malignant tumors from benign which results in accurate and speedy detection of tumor in brain along with identification of precise location of the tumor. The efficiency of the method is proved through plenty of simulations and comparisons.