ABSTRACT
Malaria is one of the deadly tropical and subtropical diseases in the world, especially in the developing countries of Africa, Asia, and Latin American continents. It is transmitted by the infected female Anopheles mosquito. Early diagnosis of malaria can reduce the mortality rate and its harmful consequences. The gold standard and widely used method of malaria diagnosis is the manual 197examination of blood smears using light microscopy. However, this method is subjective, time-consuming, and error-prone. The diagnosis result highly depends on the level of technical expertise and experience of the laboratory technicians. To improve the reliability and accuracy of this diagnosis method, automated computer-aided diagnosis (CADx) systems were proposed as a viable option. The CADx systems are used to detect the malarial parasites in the microscopic images and quantifying the level of infection. The methods proposed for detecting and classifying malarial parasites in blood smear microscopic images can be divided into two broad categories. The first methods category employs traditional image processing and classical machine learning algorithms. The second methods category employs deep learning methods for the detection and classification of malarial parasites. This work presents a comprehensive review of different methods for malarial parasite detection and classification in blood smear microscopic images. A methodological review of the recent deep learning techniques is given more emphasis. This review clearly shows the technical progresses attained in an attempt to solve this problem and future research directions.
