Automatic Leukocyte Image Segmentation: A review

Blood analysis is very important to establish the health state of a patient in order to diagnose diseases like Leukemia, one of the top causes of children mortality in Latin America, for this reason the complete blood count (CBC) is the entry examination performed by specialists. Regarding Leukemias, Leukocytes are the most studied blood components compared to Red Blood Cells and Platelets, their morphological and cytochemical properties are of ulterior relevance in the diagnosis as they unveil the development of the disease. Throughout history, the CBC has been performed manually using a microscope and specialists expertise, this is time consuming and may lead to erroneous results. Although automatic CBC hardware has been developed, it does not allow morphological visual analysis of Leukocytes so the manual method is still preferred. Image Segmentation is a procedure where an object is separated from the background for further analysis. During the last decades there has been a lot of interest in segmenting Leukocytes to automate morphological analysis in order to decrease the specialist workload and to perform faster diagnosis. In this paper a review on Leukocyte Image Segmentation, their advantages and flaws is made and a general approach towards future research is presented.