المجلة الدولية للعلوم والتقنية

In this paper, we designed and implemented a methodology to detect the mass affected by brain tumor for Magnetic Resonance Images (MRI) using MatLab program as a tool. These MRI images were collected from National Cancer Center (NCC) and Misurata Central Hospital (MCH), the total number of images used for testing stage are 50 images. All MRI images were the type of T2 weighted. The input image has been processed, which is of gray type, by changing its size to become 200*200 pixels without affecting on the image quality. Also the segmentation technique was used to extract the infected area from the image and isolate it from the rest of the healthy tissues by integrating two algorithms, Otsu algorithm to transform the image into the binary system and K-means cluster algorithm to split the image into three parts, these two algorithms gave accurate and excellent results in the detection. The accuracy rate in the detection process has reached 100% where the output image represents exactly the tumor cells that were pointed from the healthy cells.