CLASSIFICATION OF WBC USING CONVOLUTION NEURAL NETWORKS

Abstract

Leukocytes, developed within the cartilage of the bone, account for barely 1% of the overall blood cell counts. Erratic flourishing of leukocytes induces an outbreak of blood cancer. Amongst three of the diverse sorts of cancer in blood, the suggested ponder provides a vigorous instrument for the sorting of subtypes of leukemia and multiple myeloma, utilizing the related dataset. White blood cells with leukemia are not normal that grow throughout cells present in the red blood. WBCs, and platelets and affect the blood and bone marrow. Whereas, multiple myeloma is a different type of cancer that affects plasma cells. It develops in the bone marrow instead of the blood stream. The suggested method uses a deep learning technology called as convolutional neural networks to lessen the likelihood of errors occurring during the human method. The model first extracts the leading highlights from the cell imaging by pre-processing it. Next, the model will be prepared using CNN, and lastly, the cancer type can be predicted. Furthermore, the model's accuracy of 97.33% is higher than Yolov8's and Naive Bayes. Keywords: Acute Lymphoblastic Leukemia (ALL), Acute Myelogenous Leukemia (AML), Chronic Lymphocytic Leukemia (CLL), Chronic Myelocytic Leukemia (CML), Multiple Myeloma (MM), Deep Learning, CNN