Novel techniques for detection of anomalies in brain MR images

Abstract

With the significant growth in the field of medical imaging, the analysis of brain MR images is constantly evolving and challenging filed. MR Images are widely used for medical diagnosis and in numerous clinical applications. In brain MR Image study, image segmentation is mostly used for determining and visualizing the brain’s anatomical structures. The parallel research results articulated the enhancement in brain MR image segmentation by combining varied methods and techniques. Yet the precise results are not been proposed and established in the comparable researches. Thus, this work presents an analysis of accuracy for brain disorder detection using most accepted Watershed and Expectation Maximization-Gaussian Mixture Method. The bilateral filter is employed to the Watershed and Expectation Maximization-Gaussian Mixture Method to improve the image edges for better segmentation and detection of brain anomalies in MR images. The comparative performance of the Watershed and EM-GM method is also been demonstrated with the help of multiple MR image datasets.