Abstract

A new tumor classification system has been designed and developed. The magnetic resonance imaging has become a widely used method of high-quality medical imaging, especially in brain imaging where the soft-tissue contrast and non-invasiveness is a clear advantage. Recently every year human death is increased due to brain tumor. When compared with all varieties of cancer, the intentness of brain tumor is extremely big. Hence proper treatment is necessary for this. When it has unrestrained division of cells, the cerebral tumor will occur. Due to this, a typical group of cells are formed in the brain. The ordinary functioning of brain activity can be influenced by this cell group and the healthy cells get destroyed. Benign tumor and malignant tumor are the two kinds of classification in brain tumor. The benign tumor is said to be low-grade tumor. The malignant tumor is stated as high-grade tumor. Detecting the differentiation of normal brain and tumor brain is the intention of the proposed methodology. In this work, the magnetic resonance feature images are used for the tumor classification. To remove the unwanted noises in the magnetic resonance image, median filtering is used. K-means clustering algorithm is used to segment the images. Texture features are extracted using gray level co-occurrence matrix. Finally, support vector machine is used to classify the brain tumor images. It is the efficient technique and it achieves 95% of classification accuracy.

Paper Details