Leukocyte nucleus segmentation and nucleus lobe counting

摘要: Leukocytes play an important role in the human immune system. The family of leukocytes is comprised lymphocytes, monocytes, eosinophils, basophils, and neutrophils. Any infection or acute stress may increase decrease number leukocytes. An increased percentage neutrophils be caused by infection, while lymphocytes can a chronic bacterial infection. It to realize abnormal variation five types distinguished their cytoplasmic granules, staining properties size cell, proportion nuclear material, type nucleolar lobes. lobes when leukemia, nephritis, liver disease, cancer, sepsis, vitamin B12 folate deficiency occurred. Clinical neutrophil hypersegmentation has been widely used as indicator deficiency.Biomedical technologists currently recognize using eyes. However, quality efficiency diagnosis compromised due limitations biomedical technologists' eyesight, strength, medical knowledge. Therefore, development automatic leukocyte recognition system feasible necessary. essential extract region from blood smear image order develop deficiency. purpose this paper contribute nuclei segmentation method for such technology. other goal counting cell nucleus. experimental results demonstrated impressive accuracy. Insensitive variance images, LNS (Leukocyte Nuclei Segmentation) functioned well isolate with much better UR (Under Segmentation Rate), ER (Overall Error RDE (Relative Distance Error). presented LC (Lobe Counting) capable splitting into illuminated that both methods give expressive performances. In addition, three advanced processing techniques were proposed weighted Sobel operator, GDW (Gradient Direction Weight), GBPD (Genetic-based Parameter Detector).