Fuzzy Clustering based Image Segmentation techniques used to segment MRI / CT scan Brain tissues - Comparative Analysis

An analysis of image segmentation algorithms using brain tumor images

Overview

This research is a means to analyse algorithms based on Fuzzy c-means clustering that segments an image into different parts on the basis of RGB or grayscale values to be used to analyse and determine tumor cells in the brain using images from MRI/MMR scans. Currently, in order to detect brain tumor there is a higher dependency on the competency of the doctor and there is a probability of human error that could lead to incorrect or delayed diagnosis. Developing a kernel based algorithm enables it to be operable in a high dimensional dataset and makes it computationally cheaper. Thus, this research project utilises the comparative study of the existing 14 algorithms done in the past and proposes a viable computer aided detection approach to diagnose brain tumor efficiently by eliminating any error or delay caused by human fallibility or any other sources with high success rate.

Technology used

MATLAB, Fuzzy Clustering

Project link

Research paper