Efficient Brain Tumor Segmentation in Magnetic Resonance Image Using Region-Growing Combined with Level Set

Abstract

Medical image segmentation plays a great role in image processing because it can help human to extract some suspicious regions from a medical image especially brain images. Brain tumor is one of the huge medical problems. It has an influence on our lives. In this paper, we proposed a method for brain tumor segmentation and detection in magnetic resonance image (MRI). The contrast of MRI is enhanced by using histogram equalization and the tumor region is labeled by using the region-growing technique combined with the level set method to create the exact boundary of tumor region and return the segmentation result. The proposed method is better than the other recent methods based on compared results.