A Novel Technique based on Artificial Bee Colony for MR Brain Image Segmentation

Abstract

Identification of tumor present in human brain plays a crucial role in the research of medical field. Many conventional techniques have been introduced, but still there is need of accuracy in the aspect of identifying the tumor and better visualization of tissue structures located in the human brain. The suggested Artificial Bee Colony (ABC) based Fuzzy C-Means (FCM) technique performs better segmentation of tissue portions and tumor regions located in MR brain images. Feasible cluster position is obtained by the Artificial Bee Colony algorithm and clustering operation is performed by using the Fuzzy C-Means algorithm. The efficiency of the proposed technique is validated using BRATS-SICAS (2015) dataset. Segmentation result produced by proposed Artificial Bee Colony (ABC) based Fuzzy C-Means (FCM) shows improved visualization of tumor and tissue structure and it is compared with competitive segmentation techniques like Particle swarm optimization (PSO). Dice overlap index (DOI) and Sensitivity produced by the recommended algorithm are 96.2±1.8 and 97±1.4, and they are quite impressive than the competitive segmentation techniques. This novel technique can be used as an effective method to localize the tumor portion in a more accurate way than the other demarcating techniques used in the research analysis of medical images analysis.