An efficient approach for the detection of White Matter, Gray Matter, and cerebrospinal fluid from MR images of the brain using an advanced multilevel thresholding

Brain Tissue segmentation is essential in surgical planning and in diagnosing neurological diseases. This paper enlightened a novel and automatic approach for the segmentation of White Matter, Gray Matter, and CSF tissues from the MR Images of the brain. This Multilevel Segmentation of MR Images of the brain uses the concept of Harmony Search Optimization (HSO). In comparison with the other evolutionary approaches, HSO exhibits low computational complexity with interesting search capabilities. The algorithm uses the objective functions proposed by Otsu or Kapur to guide the candidate solutions, derived from an appropriate search space inside the image histogram. The operators of HSO evolve the candidate solutions until the best threshold value is found. The pursuance of the suggested method is evaluated in terms of sensitivity, specificity, segmentation accuracy using ground truth images and the results are compared with some of the existing methods. The results proved the effectiveness of the proposed approach for the MR brain image segmentation.