Malaria is a disease caused by the bite of female Anopheles mosquitoes, and is most often manifested by symptoms such as fever, chills, fatigue and vomiting. Diagnosis of malaria is still based on manual identification of Plasmodium by microscopic examination of blood cells or by Rapid Diagnostic Tests. These methods have shown their limitations, namely the need for an expert and the delay of time it takes to obtain results. This article is a review of computer-aided diagnostic systems, specifically approaches to Plasmodium detection from blood smear images. We present and compare some blood smear datasets, as well as segmentation and classification algorithms. This comparison allowed us to choose the Delgado Dataset B and Abbas et al. Dataset as dataset and YOLOv5, an object detection algorithm for our work.

Thin blood smear, Malaria disease detection, Deep learning, Image classification, Image segmentation