Multiple Cells Extraction Algorithm on Peripheral Blood Smear Images Using Modified Significance Linked Connected Component Analysis

Many diseases such as Cancer, Huntington’s disease, Malaria, Multiple Sclerosis, and Myalgic Encephalomyelitis are mainly diagnosed by the appearances include size, shape, color and the double-concave dish size of red blood cells in the blood smear. The automatically evaluating process in the appearances from the blood smear images is high importance for rapid and accurate diseases diagnosis. The proposed algorithm combines several schemes to segment red blood cells in Giemsa stained blood slides and count the appearances of red blood cells in the blood smear image for advance applications. Experiment results show that the proposed algorithm not only to provide the above appearances of red blood cells quickly and accurately, but also provides the red blood cells and double-concave dishes’ median, average, standard deviation for physicians in different areas to diagnose different diseases.