Segmentation of brain MRI using an altruistic Harris Hawks’ Optimization algorithm

Segmentation is an essential requirement in medicine when digital images are used illness diagnosis, especially, posterior tasks as analysis and disease identification. An efficient segmentation of brain Magnetic Resonance Images (MRIs) prime concern to radiologists due their poor illumination other conditions related de acquisition the images. Thresholding a popular method for that uses histogram image label different homogeneous groups pixels into classes. However, computational cost increases exponentially according number thresholds. In this paper, we perform multi-level thresholding using evolutionary metaheuristic. It improved version Harris Hawks Optimization (HHO) algorithm combines chaotic initialization concept altruism. Further, fitness assignment, use hybrid objective function where along with cross-entropy minimization, apply new entropy function, leverage weights two functions form approach. The HHO was originally designed solve numerical optimization problems. Earlier, statistical results comparisons have demonstrated provides very promising compared well-established metaheuristic techniques. article, altruism has been incorporated enhance its exploitation capabilities. We evaluate proposed over 10 benchmark from WBA database Harvard Medical School 8 Brainweb dataset some standard evaluation metrics. On dataset, Peak Signal Noise Ratio (PSNR) 26.61 Structural Similarity Index (SSIM) 0.92 achieved 5 For same scenario, PSNR 24.77 SSIM 0.86 obtained. obtained justify superiority approach existing state-of-the-art methods baseline methods. relevant codes available at: https://github.com/Rohit-Kundu/Segmentation-HHO_Altruism.