BRAIN TUMOR SEGMENTATION AND DETECTION USING MRI IMAGES

Brain tumor is caused due to the increased abnormal in brain. It is not something that we might say is limited to aged people alone, but is known to affect newborn babies as well. It affects many people worldwide. With the applications of Machine Learning (ML) and Image Processing (IP), the early detection of brain tumor is possible. In this research work, the different stages in image processing which help to detect brain tumor, is addressed vividly. This work provides information about the various sets of filtering and segmentation methods which can be used to detect whether it is brain tumor or not. All of the filtering methods are defined in image preprocessing techniques. The next procedure is to apply segmentation methods namely watershed segmentation and gray level threshold segmentation. After this, certain features are considered for feature extraction such as area, major axis, minor axis and eccentricity. According to the outcomes from the feature extraction technique, the classification of the tumor is done. In this paper, we achieve an accuracy of 92.35 by using K-Nearest Neighbor (KNN) algorithm.