Harnessing Image Analysis for Early Leukemia Detection in Bone Marrow Samples

White blood cells, or leukocytes, are important in determining a person's health since they can alter the immune system, leading to infections, anemia, leukemia, and other issues. Blood cancer is caused by these white blood cells proliferating beyond control. In this article, we develop a model for identifying and classifying cancer white blood cells using convolutional neural networks (CNNs). We train and test with the Bone Marrow Microscopic Images of Blood Samples dataset. CNN initially processes the submitted photographs. The max pooling and convolution layers then collect features, with the convolution layer handling classification and the fully connected layer handling feature extraction. The contour technique is utilized when the output indicates that the cells are cancerous. The contour approach aids in the segmentation process by allowing for the localization of the afflicted cell. The photographs in the collection are organized into four categories: neutrophils, eosinophils, monocytes, and lymphocytes. The output image shows the type of damage, as well as the mapping of the afflicted cell.