Acute Lymphoblastic Leukemia Classification with Blood Smear Microscopic Images Using Taylor-MBO based SVM

Abstract

Acute lymphoblastic leukemia (ALL) is a serious hematological neoplasis that is characterized by the development of immature and abnormal growth of lymphoblasts. However, microscopic examination of bone marrow is the only way to achieve leukemia detection. Hence, an effective leukemia detection approach is designed using the proposed Taylor-Monarch Butterfly Optimization based Support Vector Machine (Taylor-MBO based SVM). However, the proposed Taylor-MBO is designed by the integration of Taylor series and Monarch Butterfly Optimization (MBO), respectively. The sparking process is designed to perform automatic segmentation of blood smear image by estimating optimal threshold values. By extracting the features, such as texture features, statistical and grid-based features from the segmented smear image, the performance of classification is increased with less training time However, the proposed Taylor-MBO based SVM obtained better performance using the metrics, such as accuracy, sensitivity, and specificity with the values of 94.5751%, 95.526%, and 94.570%, respectively.