Fuzzy Based Blood Image Segmentation for Automated Leukemia Detection

Acute lymphoblastic leukemia (ALL) are a group of hematological neoplasia of childhood which is characterized by a large number of lymphoid blasts in the blood stream. ALL makes around 80% of childhood leukemia and it mostly occur in the age group of 3-7. The nonspecific nature of the signs and symptoms of ALL often leads to wrong diagnosis. Diagnostic confusion is also posed due to imitation of similar signs by other disorders. Careful microscopic examination of stained blood smear or bone marrow aspirate is the only way to effective diagnosis of leukemia. Techniques such as fluorescence in situ hybridization (FISH), immunophenotyping, cytogenetic analysis and cytochemistry are also employed for specific leukemia detection. The need for automation of leukemia detection arises since the above specific tests are time consuming and costly. Morphological analysis of blood slides are influenced by factors such as hematologists experience and tiredness, resulting in non standardized reports. A low cost and efficient solution is to use image analysis for quantitative examination of stained blood microscopic images for leukemia detection. A fuzzy based two stage color segmentation strategy is employed for segregating leukocytes or white blood cells (WBC) from other blood components. Discriminative features i.e. nucleus shape, texture are used for final detection of leukemia. In the present paper two novel shape features i.e., Hausdorff Dimension and contour signature is implemented for classifying a lymphocytic cell nucleus. Support Vector Machine (SVM) is employed for classification. A total of 108 blood smear images were considered for feature extraction and final performance evaluation is validated with the results of a hematologist.