Automated segmentation of acute leukemia using blood and bone marrow smear images: a systematic review.

Acute leukemia is a proliferation of white blood cells that originates in the bone marrow and impinges on the working of the lymphatic system. Due to the abrupt and aggressive nature of spreading, it requires prompt medical attention. The procedure adopted by hematologists for evaluating morphological features and counting the cells is quite time-consuming, cumbersome, and error-prone. Thus, automatic methods are recommended to overcome the limitations of manual segmentation. The motive of this systematic review is to investigate the various automatic segmentation techniques of acute leukemia using blood and bone marrow smear images. The guidelines of PRISMA have been adopted for systematic review. For the identification of the papers, three online databases are used, namely PubMed, Science Direct, and Google Scholar. A query has been used to find the relevant research article. Inclusion and exclusion criteria are used to scrutinize the papers, and four research questions have been framed to explore the literature to investigate further. In the preview of four different questions, the various methods have been compared based on accuracy, and the various challenges have been highlighted with respect to the methods and distinct datasets of blood and bone marrow smear images. This systematic review summarizes the various automated segmentation techniques under the pretext of framed research questions based on the PRISMA Model. The different types of datasets, performance evaluation parameters, and various obstacles with future scope have also been discussed. [ABSTRACT FROM AUTHOR]