Identification of Calcification in MRI Brain Images by k-Means Algorithm

Background/Objective: The role of clustering is significant to analyze different kind of applications of its techniques. Similar data are grouped into one and they formed as a cluster. Dissimilar data are grouped into another form in other cluster. Data clustering is an important and active research applied in many fields including multivariate analysis in statistics and some other areas like pattern recognition and machine learning etc. Methods/Statistical Analysis: Boundary detection and outlier analysis is an important concept for pre-processing the data. The boundary considers only pixels lying on and near edges and use of gradient or Laplacian to preliminary processing of images. To find the outlier in a group of patterns is a well-known problem in Data Mining (DM). An outlier is a pattern which is different with respect to the rest of the patterns in the data. The k-Means is one of the familiar clustering methods used by different researchers to find the well-formed clusters. Magnetic Resonance Imaging (MRI) uses a magnetic field and radio waves to create detailed images of the organs and tissues within human body. The k-Means algorithm is used to find the tumor by applying the boundary detection and outlier techniques in this research work in MRI brain images. Findings: The main goal of this research work is to extract the tumor (Calcification) in an MRI brain image by means of clustering pixels to fortify the quality of clustering algorithm. The results of the MRI brain images are analyzed and identified by the proposed algorithm. The result produced by simple k-Means algorithm is very useful to find the tumor in MRI images perfectly. Application/Improvements: The MRI brain images are analyzed and implemented by other methods like classification and some other techniques in future.