期刊名称：International Journal of Computer Science and Information Technologies
电子版ISSN：0975-9646
出版年度：2015
卷号：6
期号：3
页码：2593-2596
出版社：TechScience Publications
摘要：Detection of tumors such as normal, benign and malignant stage tumors are detected. This method includes MRI brain image classification and segmentation approach. It is an learning machine approach where the network is trained with PNN-RBF and it is used to detect the tumor regions along with the spatial fuzzy clustering method. The classifier includes wavelet transform followed by feature extraction called GLCM and the training network called PNN-RBF. Finding benign stage tumors improves the chances saving the human life. The proposed segmentation process includes Spatial fuzzy clustering where the process is carried out by calculating centroid, distance and membership. The neural network is trained with images in each benign, malignant and normal stage. The system also uses to segment the tumor cells along with this morphological filtering which is used to remove background noises for smoothening up of region. The result includes segmented tumors with parameter evaluation. The segmentation prcess also results with the number of cells defected and the measurement of tumor area detected.
