Contrast enhanced brain tumor segmentation based on Shannon's entropy and active contour

In this paper, a novel computer procedure is proposed to assist the brain tumor image examination. This approach enhances and extracts the contrast improved tumor core section from a two dimensional Magnetic Resonance Image (MRI) integrating the Bat Algorithm(BA), Shannon's multi-thresholding, and Active Contour (AC) based segmentation. Firstly, BA assisted multi-thresholding is executed to improve the tumor core section of the brain MRI dataset. Later, the tumor core is extracted using the AC segmentation approach. The proposed methodology is tested on the well-known BraTS MRI dataset. The success and the clinical significance of the proposed approach are verified using image similarity values and statistical measures. The experimental results confirm that the proposed approach presents a great performance when compared with the ground truth, suggesting that it might have real world practical implications with a clinically significant impact.