An Automatic Brain Tumor Segmentation and Classification using MRI Image

Arbaz Khan, Chidananda Murthy, M Kurian
International Journal of Biotechnology and Biomedical Sciences   unpublished
The aim of the project is to detect and classify the brain tumor from MRI image. This project involves mainly 6 stages namely Input Image, Preprocessing, Segmentation, Post Processing, Feature Extraction and Classification. In this phase, 4 stages are implemented, Input image, preprocessing, segmentation and post processing. Input image reads the MRI brain image. Preprocessing mainly includes image smoothing and image enhancement. Image smoothing can be accomplished by using Median Filter and
more » ... Median Filter and which is followed by Image Enhancement technique which can be achieved by Sobel edge detect technique. The third stage is segmentation. In this project, brain tumor is segmented using Pillar K-Means algorithm. Pillar K-Means algorithm includes selection of pillar pixels for useful segmentation. The experimental results shows that the proposed algorithm can effectively segments the tumors from MRI. Post processing operations are applied on the image to clearly locate the tumor part in the brain.