Segmentation and Classification of White Blood SMEAR Images using Modified CNN Architecture

Abstract

The classification and recognition of leukocytes or WBCs in blood stain images present a key role in the corresponding diagnosis of specific diseases, such as leukemia, tumor, hematological disorders, etc. The segmentation & classification of WBCs nucleus sets a solid ground for automation in the calculation and recognition of WBCs related disorders. However, it is complicated by reason to causes such as background obscurity and alterations in occurrence generated by histological staining conditions. For improving the accuracy of segmentation of WBC images, we intend to consider a deep learning network using microscopic images of white blood cell. Furthermore, the functioning of the intended design will be compared with other modern methods. By the end of this study a successful model classifying various nucleus morphologies such as Basophil, Eosinophil, Lymphocyte, Monocyte and Neutrophil was obtained where overall test accuracy achieved was 97.0 % for VGGNet classifier and 94.0 % for MobileNet V2 classifier.