A Deep Learning Approach for Blood Cells Classification Using CNN and Transfer Learning Models

Blood cell image classification stands as a cornerstone in modern healthcare, playing a pivotal role in the diagnosis and treatment of a wide array of medical conditions. The precise and efficient categorization of blood cells holds the potential to exert a profound influence on patient care and medical research alike. The traditional methods faced challenges with accuracy and robustness due to the complexity and variability of blood cell images. This work, propose a comprehensive approach for the classification of four types of blood cells: Eosinophils, Lymphocytes, Monocytes, and Neutrophils using Convolutional Neural Networks (CNNs) and transfer learning models. We employ four state-of-the-art pre-trained models: Xception, VGG16, MobileNetV2, and ResNet50V2. The primary objective is to accurately classify blood cell types, crucial for various medical diagnostic applications. Our experiments show's promising results, with the MobileNetV2 model achieved an impressive Training Accuracy of 98.22%, Validation Accuracy 86.89%, and Test Accuracy of 87.00%.