Recognition System for Malarial Parasites Causing Protozoa Infections in Thin Blood Smears

Fungi, bacteria, and protozoa are the primary agents of infectious diseases. Bacteria are the prokaryote kingdom, and yeast mold belongs to the fungi kingdom. Protists are the kingdom of protozoa. The malaria parasite is nothing but Plasmodium, a genus of parasitic protozoa. Parasites are organisms that use another organism directly to their benefit. This research article uses a computer-aided diagnostic tool to recognize malarial parasites that result in protozoa infections in thin blood smears. Six different machine learning algorithms were used, out of which the support vector machine resulted in 89.94% accuracy. The results are validated using the proposed Convolution Neural Network (CNN) model. With convolution and pooling processes for each pixel, CNN learns more compelling features, thus resulting in a much better decision hyperplane. The different layers were designed to get better accuracy. The accuracy of 96.46% is attained using CNN, which results in a 6.76 % accuracy improvement.