FUZZY IMAGE SEGMENTATION USING ANT COLONY OPTIMIZATION

Abstract

Segmentation plays a vital role in medical image processing. Effective segmentation is necessary for structural analysis of an organ, diagnosis and detection of certain abnormalities. Magnetic Resonance Imaging (MRI) is an advanced technique used in field of medical imaging. Manual image segmentation is very tedious and time consuming. Also results of manual segmentation are subjected to errors due to huge and varying data. Therefore, automated segmentation systems are gaining enormous importance nowadays. This study presents an automated system for segmentation of brain tissues from brain MRI images. Segmentation of three main brain tissues is carried out namely white matter, gray matter and cerebrospinal fluid. In this work, we performed the initialization step for fuzzy C-means clustering algorithm using Ant Colony Optimization. Clustering results are often dependent upon the initial solution. ACO is a meta-heuristic approach inspired by the intelligent behaviour of real ants which provides close to optimal solution avoiding any trap in local minima. Spatial information is also considered in segmenting brain tissues as grouping of pixels into different clusters is influenced by its local neighbourhood. Also Mahalanobis distance metric is used instead of Euclidean distance metric in clustering process to avoid any relative dependency upon the geometrical shapes of different clustering classes. The results of the system are evaluated and validated against the ground truth images for both real and simulated database.