Classification of Leukemia Microscopic Images using Blended Biogeography Optimization

The need for accurate medical images classification technique for assisting the clinician in disease diagnosis is increasing day by day. This research paper investigates the performance of blended biogeography based optimization for Acute Lymphoblastic Leukemia classification and the input microscopic images are categorized as abnormal and normal. In this analysis, 47 cancer cells and 26 healthy cells taken from C- NMC website are considered. The better performance of blended biogeography based optimization is established through comparison with original biogeography based optimization, particle swarm optimization, K-Means and Fuzzy C means techniques. The proposed technique yields highest accuracy of 93% when statistical features are used.