Classification of WBC Using VGG-19 Architecture

Abstract

White Blood Cell (WBC) classification plays a crucial role in diagnosing various hematological disorders and infections. Automated WBC classification using deep learning techniques has gained significant attention due to its efficiency and accuracy. This conference paper presents a study on the application of the VGG-19 deep learning model for WBC classification. The research highlights the methodology, experimental results, advantages, challenges, and future research directions in utilizing VGG-19 for medical image analysis, particularly in hematology. White Blood Cell (WBC) classification plays a crucial role in diagnosing various hematological disorders, including leukemia and infections. Traditional manual methods of WBC classification are time-consuming and prone to human error. Deep learning techniques, particularly Convolutional Neural Networks (CNNs), have demonstrated remarkable accuracy in medical image classification. This study explores the use of the VGG-19 architecture for automated WBC classification to enhance diagnostic efficiency and accuracy. The proposed model is trained on a labeled dataset of WBC images, leveraging the deep hierarchical features of VGG-19 to classify different WBC types, such as neutrophils, eosinophils, basophils, monocytes, and lymphocytes. Transfer learning is employed to fine-tune the pre-trained VGG-19 model, improving its performance on the specialized medical dataset. The model's performance is evaluated using accuracy, precision, recall, and F1-score metrics. Experimental results indicate that the VGG-19-based approach achieves high classification accuracy, outperforming conventional machine learning methods.