Multilevel Thresholding of Brain Tumor MRI Images: Patch-Levy Bees Algorithm versus Harmony Search Algorithm

Abstract

Image segmentation of brain magnetic resonance imaging (MRI) plays a crucial role among radiologists in terms of diagnosing brain disease. Parts of the brain such as white matter, gray matter and cerebrospinal fluids (CFS), have to be clearly determined by the radiologist during the process of brain abnormalities detection. Manual segmentation is grueling and may be prone to error, which can in turn affect the result of the diagnosis. Nature inspired metaheuristic algorithms such as Harmony Search (HS), which was successfully applied in multilevel thresholding for brain tumor segmentation instead of the Patch-Levy Bees algorithm (PLBA). Even though the PLBA is one powerful multilevel thresholding, it has not been applied to brain tumor segmentation. This paper focuses on a comparative study of the PLBA and HS for brain tumor segmentation. The test dataset consisting of nine images was collected from the Tuanku Muhriz UKM Hospital (HCTM). As for the result, it shows that the PLBA has significantly outperformed HS. The performance of both algorithms is evaluated in terms of solution quality and stability.