Performance Study of Various Image Segmentation Techniques for Brain Tumor Detection

Abstract

The detection of tumor in brain from Magnetic Resonance Image (MRI) is a very challenging and difficult task performed by clinical experts and it is made easier through computer technology called medical image segmentation. This technology partitions the given brain MRI image into number of segments to obtain desired location of tumor. The technology not only helps in detecting tumors in early stages but also contributes towards treatment based on the analysis thus saving thousands of lives.  This paper discusses about various image segmentation techniques such as Edge based, Region based and Pixel based to detect tumors in MRI brain image. Performance of various techniques were tested using MATLAB for tumor detection is compared based on various parameters.