USING CONVOLUTIONAL NEURAL NETWORKS FOR FINE GRAINED IMAGECLASSIFICATION OF ACUTE LYMPHOBLASTIC LEUKEMIA

Acute lymphoblastic leukemia (ALL) is a cancer of bone marrow stems cells that results in the overproduction of lymphoblasts. ALL is diagnosed through a series of tests which includes the minimally invasive microscopic examination of a stained peripheral blood smear. During examination, lymphocytes and other white blood cells (WBCs) are distinguished from abnormal lymphoblasts through fine-grained distinctions in morphology. Manual microscopy is a slow process with variable accuracy that depends on the laboratorian\u27s skill level. Thus automating microscopy is a goal in cell biology. Current methods involve hand-selecting features from cell images for input to a variety of standard machine learning classi ers. Underrepresented in WBC classi cation, yet successful in practice, is the convolutional neural network (CNN) that learns features from whole image input. Recently, CNNs are contending with humans in large scale and ne-grained image classi cation of common objects. In light of their e ectiveness, CNNs should be a consideration in cell biology. This work compares the performance of a CNN with standard classi ers to determine the validity of using whole cell images rather than hand-selected features for ALL classification