Sickle Cell Anaemia Detection using Deep Learning

Document Type : Original Article

Authors

1 computer science, faculty of computers and artificial intelligence, fayoum university

2 Bioinformatics, faculty of computers and artificial intelligence, fayoum university

Abstract

Many disorders, including sickle cell anaemia which causes periodic bouts of pain and severe, pronounced anaemia, result in red blood cell (RBC) deformation. It takes longer to monitor patients with these disorders since peripheral blood samples must be examined under a microscope. The observation of isolated RBCs is subjective; hence the error rate is considerable and an expert is needed to perform this approach, SCD can be adequately managed and the death rate can be decreased with early detection. Therefore, this work proposes a deep learning method for sickle cell detection based on Convolutional neural networks (CNN). VGG model differentiate between three classes of red blood cells which are circular (normal), elongated (sickle cells), and other blood content. it is applied on ERYTHROCYTESIDB dataset for validation. A comparison of the results showed that the proposed model is superior for the diagnosis of sickle cell anaemia with 99.4% of overall accuracy.

Keywords

Main Subjects

Files

Share

How to cite

Statistics