The Use of YOLOv5 as a Malaria Detection Model for the Developing World

This study addresses the challenges associated with malaria diagnosis, particularly in regions with high disease prevalence and high resource constrained. The research leverages advanced biomedical technologies, including microscopes and image processing techniques, to enhance the identification of malaria parasites. We study the effectiveness of YOLOv5, an object detection model, for malaria detection. The model’s performance is assessed using a real dataset comprising of images of blood smears infected with malaria parasites. The evaluation considers various parasite types in both their original and processed image forms. The presence of staining variation and differences in color settings across slide images posed challenges, impacting the model’s performance on preprocessed images. Results show that the studied model has the potential to significantly improve early and accurate malaria diagnosis, contributing to more efficient disease management and control strategies.