Detection and classification of white blood cell cancers

Abstract

Technology in this era is moving exponentially high. Adoption of technology in almost every aspect of our lives at all times can be found easily. To detect white blood cells cancer by image processing is one among approach. An automatic and novel approach for the analysis of white platelets diseases such as Acute lymphoblastic leukemia (ALL), Acute myeloid leukemia (AML), Chronic lymphocytic leukemia (CLL) and Chronic myeloid leukemia (CML) is proposed which is a challenging biomedical research topic. Division of disorder into various categories; every classification incorporates comparable side effects that confound in diagnosing. Various signs includes shading, size, shape and structure and a particular methodology is applied by computing different features. The main purpose is to remove the time delay to get the results from the Existing systems. The proposed plan is based on pre-processing via inbuilt MATLAB tools, segmentation of white blood cells using Watershed algorithm, feature extraction by working on Grey Level Co-occurance(GLCM) and classification by testing Support Vector Machine(SVM). Pre-defined data set is used, but in the future, live data set can also be used for the same by using a biometric-sensor which takes microscopic images of the blood in real time. The precise experiment of the proposed methodology was conducted and has significantly excelled the existing methodology by overcoming time delay and slow access to more reliable and easy implementation in respect to white blood cell’s leukaemia classification.