Detection of Brain Tumor Region in MRI Image Through K-Means Clustering Algorithms

The brain is the primary central control area of the human nervous system. The growth of tumors is a quick and abandoned development of abnormal brain cells. The approach often used to classify the tumor area within the brain is magnetic resonance imaging (MRI). MRI pictures are typically not that well suited for extensive examination since, in this epoch of current medical technology, the study of medical images is one of the most critical, challenging, and growing fields. These suggested techniques are used to emphasize tumor boundaries. The research technique is focused on filtration, border preservation, and segmentation of three key tasks. For this reason, it is the first step to eliminate the noise from the MRI images of salt and pepper by using a median filter for efficiency purposes, and to retain the edges such that the tumor area is increased. To build the hybrid edge operator and view all entities present on the MRI image, the two algorithms used in this paper are K-means and DPSO algorithms. Segmentation is one of the most effective strategies for measuring and diagnosing brain tumors. To separate images in many areas, segmentation techniques are used. In segmenting pixels several different approaches, such as C-means clustering, K-means clustering, watershed segmentation, etc., are labeled with the same characteristics. This paper demonstrates the difference between all approaches at each point and also the hybridized outcome is obtained in the MRI image by extracting tumor area.