Examination of the Brain MRI Slices Corrupted with Induced Noise—A Study with SGO Algorithm

Abstract

Brain irregularity is a chief disease in humans and conscientious for the loss of many lives. Due to its consequence, several image appraisal methods are proposed to study the abnormality in the brain. This research implements a two-stage practice to extort the cancer fragment from axial-view brain MRI slice stained with noise. Primarily, the selected 2D MRI slice is stained with the regular noises, and then the threshold and segmentation tactics are applied to mine the tumor sector. In this work, brain MRIs of Flair/T2 modalities are considered, and the thresholding is executed using Social Group Optimization (SGO) and Tsallis Entropy (SGO + TE) to improve its condition and later the Level Set Segmentation (LSS) is applied to extract the tumor. Lastly, the tumor area is compared against the Ground Truth (GT) to substantiate the benefit of the projected procedure. The outcome of this work verifies that, implemented work supplies improved result in original and noise corrupted MRI slices. The execution of this procedure is also quite straightforward compared with other procedures, so this work can be used to inspect the clinical MRIs in the future.