Blood Cell Image Classification Using Deep Sequential Model

Abstract

Blood cell classification plays a vital role in medical diagnostics, aiding in disease detection and treatment monitoring. Traditional methods of blood cell analysis are often manual, time-intensive, and prone to human error. Deep learning, specifically sequential models, offers a promising solution to automate and enhance accuracy in blood cell classification tasks. This study aims to develop and evaluate a deep sequential model for classifying blood cell images into four types: Eosinophil, Lymphocyte, Monocyte, and Neutrophil. The deep sequential model was trained on a publicly available blood cell dataset. The dataset underwent preprocessing, including resizing, normalization, and augmentation. The model architecture included convolutional layers for feature extraction, max-pooling for dimensionality reduction, and fully connected layers for classification. Model performance was evaluated using metrics such as accuracy, precision, recall, and F1-score. The model achieved an overall accuracy of 99.7%, with high precision, recall, and F1-scores across all classes. The confusion matrix indicated robust classification, with minimal misclassifications. This study demonstrates the potential of deep sequential models in automating blood cell classification with high accuracy and reliability, offering a scalable solution for hematological diagnostics.