Classification of Brain Tumor Using Texture Features with Segmentation & FFNN Approach

File Size:
1.00 MB
Volume 3, Issue 8 (August, 2017)
Publication No:
Parul Verma, Jasmeet Singh Gurm
4 x

Computer aided technology makes various advancements in medical image process. In this paper a novel approach has been proposed for segmentation of brain tumor images and classification of tumor and non-tumor images from a regular dataset. Texture feature have been computed for dataset that has been used as input for feed forward neural network that having number of hidden layers. On the basis of these features classification has been done and threshold based algorithm has been used for extraction of tumor effected region that has been used for computation are and perimeter of tumor so that decision can be taken

Brain Tumour, segmentation, Neural Network, GLCM and ROC