Recently, Medical imaging is the process by which a physician can examine the inside of a patient’s body without operating it through the development of devices for the diagnosis and treatment of patients. One intriguing medical imaging method that is thought to be a highly helpful tool for identifying the tumor growth of multiple sclerosis is magnetic resonance imaging (MRI). An efficient method for separating brain tumors from MRI pictures is segmentation. For the automatic identification of MS outliers, a number of recent methods for the segmentation and classification of MRI sequences have been put forth. The “Ant Colonies Optimization ACO” meta-heuristic technique is presented in this paper for MRI image segmentation. In order to optimize their overall rendering and compare with consensual segmentation, we suggest using the ant colony technique to estimate the segmentations of brain MRI images from a public MR dataset of 30 MS patients that were imaged with a 3T MR scanner using conventional sequences. We used MATLAB GUI program to evaluate the performance of this algorithm optimized by maximizing the intra-class variance criterion MVar.