Enhanced leucocyte classification for leukaemia detection

Abstract

The measure of blood cells are more important for the doctor to diagnose various diseases such as anaemia, leukaemia etc. Similarly observation and classification of these cells grant for the estimation and detecting of a vast number of sickness. By evaluating white blood cells (WBCs) allows the leukaemia detection (Acute Lympho blastic leukaemia (ALL), Acute Myloid leukaemia),be cancer which affected on blood which can be lethal if it can be untreated. So precise counting and classification of blood cells have an important role. More over the counts especially differential counts and shape provide a lot of valuable information to evaluate leukaemia. In current methods the morphological evaluation of haemocytes is performed manually by experts and counting of blood cells is done using a device called Haemocytometer. Yet, this approach has so many limitations, such as slow estimation, different standard, deviant accuracy and dependence on the operator's skill. For counting hardware solutions such as the Automated Haematology Counter exists, they are very expensive, unaffordable in every hospital laboratory and also use actual blood samples. So there consistently need a cost effective, simple and robust method for counting,analysis and classification of blood cells. The proposed method describes a complete automatic computerized method for WBC identification,counting and classification using microscopic images.