Computer Aided Diagnostic System for Detection of Leukemia Using Microscopic Images

In present scenario, hematological disorders of leukocyte (WBC) are very frequent in medical practices. This work proposes a novel technique to differentiate ALL (acute lymphoblastic leukemia) lymphoblast cells from healthy lymphocytes. The technique first separate leukocytes from the other blood cells and then lymphocytes are extracted. In this context, a novel computer aided diagnostic system (CAD) is designed for detection of hematological disorders like leukemia (blood cancer) based on Gray level co–occurrence matrices (GLCM) and shape based features. The features thus extracted classified by the auto support vector machine (SVM) binary classifier to find the presence of lymphoblast cell (leukemic cells). GLCM texture feature with feature vector length 13 reveals, classification accuracies of 86.7% and 72.4% for cytoplasm and nucleus respectively while for shape based features illustrated, classification accuracies of 56.1% and 72.4% respectively for a feature vector length 11 in both regions of lymphocyte. The classification accuracy of combined texture-shape feature is 89.8% with feature vector length 37 which shows better results as compared to an individual.