A New Hybrid Adaptive Cuckoo Search-Squirrel Search Algorithm for Brain MR Image Analysis

Abstract

This chapter presents a new hybrid adaptive cuckoo search-squirrel search (ACS-SS) algorithm for brain magnetic resonance (MR) image analysis. Thresholding is one of the popular methods utilized for brain image segmentation. Thresholding-based methods are easily implemented. In this context, we present an optimal multilevel thresholding technique for brain MR images using edge magnitude information. The edge magnitude is computed using the gray-level co-occurrence matrix (GLCM) of the brain image slice. The optimum thresholds are found by maximizing the edge magnitude. A new hybrid evolutionary computing technique, namely ACS-SS, is investigated to maximize the edge magnitudes. The proposed scheme is tested with T2-w brain MR images from Harvard medical education database. The results are compared with cuckoo search (CS), squirrel search (SS), and adaptive cuckoo search (ACS) algorithms. It is witnessed that the findings, using the proposed ACS-SS technique, are superior to the other techniques in terms of qualitative and quantitative measures. The advantages of the proposed technique are as follows: (i) The ACS-SS shows improved fitness function values; (ii) the ACS technique gives speed improvement.