Comparative Study on Segmentation Techniques for Biomedical Images

Abstract

Segmentation is one of the most useful pillars in the medical image processing field, especially for tumors diagnosis and early detection. It is the process of partitioning the image into different regions to extract the object of interest which is the tumor in this work. This paper presents a comparison between different segmentation techniques applied on brain tumor magnetic resonance imaging (MRI) images, as a case study. The techniques under comparison are Region-Growing, Active-Contour, Graph-Cut and Global Thresholding. The performance of these techniques is evaluated based on the Jaccard Index, Dice Index and the F-score, elaborating which one is more accurate than the other.