Brain tumor segmentation using threshold method

Abstract

Tumor cells are uncontrolled cells that grow uniformly and without control when they are deprived of normal factors. The death rate from brain tumors among humans is high every year. In the United States, about 50 patients diagnosed with primary brain tumors die each year. Magnetic Resonance Imaging (MRI) is one of the most commonly used and popular methods of diagnosing brain tumors. This research work presented an automated method for brain tumor detection using gray-scale MRI images. This method involved initial enhancement to reduce gray-scale colour fluctuations. A filter was used to improve segmentation.A threshold-based OTSU segmentation instead of colour segmentation in this study because the images were grayscale. Finally, pathology experts gave information that was used to establish the study's focus areas (brain tumor region). The testing results revealed that the proposed strategy outperformed existing available alternatives in terms of accuracy while retaining an acceptable accuracy rate for pathology experts.