Manuscript received August 1, 2025; revised August 22, 2025; accepted October 27, 2025; published March 26, 2026.

Abstract—Microscopic molecules, such as White Blood Cells (WBCs) or leukocytes, are used in somatic immunity. The classification of white blood cells into neutrophils, basophils, eosinophils, lymphocytes, and monocytes has become very important. With the development of artificial intelligence, which has become a powerful tool for enhancing efficiency and accuracy in the medical system, it has improved the quality of healthcare by supporting physicians in making immediate clinical decisions based on comprehensive evidence and knowledge. In this paper, a Normalized Deep Learning Model (NDLM) has been proposed to classify white blood cell types which can be considered as a developed Convolutional Neural Network (CNN), and it is a type of deep learning. It has a specific structure because the goal of our model is to classify white blood cells, and the cells are very precise. Therefore, we need small filters to capture the fine details of the cells. While medical images require fine filters, unlike nature images. The leukocyte types were classified with the result of a high accuracy of 96.19% by the suggested model in this paper for five types of white blood cells: neutrophils, basophils, eosinophils, lymphocytes, and monocytes. The second step of this paper is the comparison between the results of this paper and other papers, where the suggested model attained superior performance.

Keywords—Convolutional Neural Network (CNN), deep learning, White Blood Cells (WBCs), classifier

Cite: Ali Rakan Hasan Al-Jader, Neam Salim M. Sheet, Mazin N. Farhan, and Raid Rafi Omar Al-Nima, "Normalized Deep Learning Model for Classifying White Blood Cell," Journal of Image and Graphics, Vol. 14, No. 2, pp. 259-266, 2026.

Copyright © 2026 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC-BY-4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.