MR Brain Image Segmentation Optimized by Using Ant Colony Algorithm with BrainSeg3D for Multiple Sclerosis Tumors Detection

Abstract

Nowadays, Brain tumors segmentation is a very important task in diagnosis of brain lesions.There are diversapproaches of segmentation, but segmentation of brain tumors is always difficult due to the complex features of magnetic resonance imaging MRI such as the appearance of vague and unclear tumors and boundaries. In this paper, an automatic MRI segmentation method is used to solve these problems. Here, tumor segmentation is treated as a problem of classification using the Ant Colony ACO optimization algorithm combined with a proposed protocol based on BrainSeg3D tools. Many studies and many existing approaches tend the multiple sclerosis (MS) which is a chronic inflammatory anomaly of the central nervous system. To pick up the outliers of multiple sclerosis, we should do a diagnosis by magnetic resonance imaging with various sequences. For this aim, some late technics of segmentation and classification of MRI images have been suggested for the automatic detection of MS outliers. In this work, we submit a novel protocol based on current and novel semi-automated tumors segmentation technics of BrainSeg3D. Evaluation of our results was performed on novel MR database containing 30 MS patients, which were learned with a 3T MR scanner with conventional sequences so that to estimate our assessment with the acquired consensus segmentation with ground truth data. The submitted algorithm is evaluated by using MATLAB GUI program and BrainSeg3D tools.