Brain tumor is an abnormal cell population that occurs in the human
brain. Nowadays, medical imaging techniques play an essential role in tumor
diagnosis. Magnetic resonance imaging (MRI) is a medical imaging technique
that uses radio waves and a magnetic field as sound waves are created to produce
detailed images of tissues and organs in the human body by computer. In this study,
three different methods were reviewed and compared to the tumor’s extraction
from a set of MRI brain images. These methods are seeded region growing, kmeans,
and global thresholding. The images used in this study are obtained from
the Cancer Imaging Archive (TCIA) and Kaggle. All images are grayscale and in
JPEG format.The images fromTCIA dataset are 100 images that contain abnormal
(with a tumor) brain MRI images while there are 35 images in the Kaggle dataset.
The Kaggle dataset contains 20 normal images and 15 abnormal images. The
results show that the k-means segmentation algorithm performed better than the
others on TCIA dataset according to the Root Mean Square Error (RMSE), the
Peak to Signal Noise Ration (PSNR), and Segmentation Accuracy while global
thresholding is the best on Kaggle dataset.