Abstract: In recent years, technical advances and enhancements in Magnetic resonance imaging (MRI) have aggravated increasing interest in the clinical role. The Computer Aided Diagnosis (CAD) framework helps the radiologists in researching the irregularities and helps radiologists to distinguish inconspicuous variations from the normal to the more critically, potential brain diseases at the prior stages. The CAD tool acts like a checker for therapeutic image and helps radiologists by highlighting regions that warrant a second examination. The development of computer aided detection technology which laid foundation for solving early diagnosis of brain tumor. The CAD tool for medical image processing is proposed in this paper. The tool is developed in MATLAB to read images of different formats like tif, jpg, DICOM etc. The tool fits for showing data about the loaded image of selected format, read and spare images form and to work space. This paper also aims at extracting the meaningful objects lying in the image and goes for extricating significant objects in the image In this paper, K-Region based Clustering and watershed segmentation has been proposed. K-means is one of the prominent methods of its simplicity and computational efficiency, whereas watershed segmentation is one such dependable way to deal with homogeneous regions in the image. It is simple, can be parallelized and dependably produces a complete division of the image. The four parameters, such as Probabilistic Rand Index (PRI), Variation of Information (VOI), Global Consistency Error (GCE), Peak signal to noise ratio (PSNR), and Structural similarity (SSIM) have been utilized to evaluate the performance. The performance is evaluated by taking the Brain MRI images as the input and GUI based CAD tool has been used to evaluate the performance.

Keywords: Brain MRI image, computer Aided Diagnosis, Image segmentation, k-means, Watershed segmentation.