A Fast Segmentation Method for the Recognition of Acute Lymphoblastic Leukemia using Thresholding Algorithm

In medical diagnosis system, vast number of diseases are diagnosed on the basis of counting and classification of blood cells. Acute lymphoblastic leukemia (ALL) is the most common type of blood cancer of white blood cells (WBCs) in children below 7-8 years. It can be fatal if diagnosed late or untreated. ALL cells are abnormal lymphocytes called blast cells or lymphoblast. Careful microscopic examination of stained blood smear or bone marrow aspirate is the best way to diagnose leukemia. There are various techniques such as cytochemistry, FISHFluorescence In Situ Hybridization, cytogenetic analysis etc. are available for specific leukemia detection. These tests are done manually so time consuming as well as costly. Therefore low cost and efficient solution is automatic detection and analysis of microscopic blood images. This paper presents complete and fully automatic method for WBCs identification and classification of blasts from microscopic images. The proposed method is to segment normal and ALL lymphocytes into two parts: nucleus and cytoplasm. This is done by using Otsu’s thresholding algorithm. The MATLAB is used to develop the whole work.