Development of AI based Techniques for Brain MR Image Segmentation

Abstract

Brain magnetic resonance (MR) image segmentation is an interdisciplinary newlinefield at the intersection of engineering and medicine. This field is intended newlineto develop computational and mathematical models for solving problems newlinepertaining to medical images and their use for biomedical research and clinical newlinecare. The main goal of the thesis is to delineate different tissue region in the brain newlineMR images. The segmented tissue regions may assist the image based diagnosis newlineand provide clinically relevant information. The thesis focuses on the retrospective newlineprocessing of the brain MR images. Every year, many more articles are added in newlinethe literature that describe new algorithms in solving the problem. However, the newlineartifacts such as intensity inhomogeneity (IIH), noise, inaccurate tissue boundaries, newlineetc. make the brain MR image segmentation a challenging task. It still remains an newlineopen problem. In this thesis, we put an effort to overcome the challenges using AI newlinebased techniques. newline