Feature Fusion based Deep Learning method for Leukemia cell classification

Pathologists typically diagnose leukemia by looking at a blood smear under a microscope. Leukemia may cause due to leukocytes abnormal proliferation in blood and bone marrow. It may also indicated by decrease in numbers of other cells combined with immature leukocytes. The types of leukemia is confirmed by examination of bone marrow by the pathologist. The traditional method for identification of types of leukemia cells is time consuming and accuracy depends upon the expertise of the professional. Hence, an automated method require which can reduce the time and identify cells with high accuracy. In this regards, several machine learning, deep learning techniques are available which take image as an input and predict the types of leukemia cells. In the proposed work a feature fusion, based deep learning method has been utilize to classify blood smears. The fusion-based model can classify with an accuracy of 99.3%.