Automatic classification of Malaria Using Artificial Neural Network

Abstract

Malaria is one of the three most serious diseases worldwide, affecting 225 million infections each year in Sudan, mainly in the tropics where the most serious illnesses are caused by Plasmodium parasite. Automatic diagnosis design systems have been implemented to detect the presence of two types of Malaria (falciparum, vivax) using neural network. Firstly, the data has been acquisition from website and laboratory. The images were filtered to remove the noise using morphological filter, in order to separate the parasite from the other cell in the image k-means method is carried out. Then features (statics first order) were selected from textural features by t-test method, and neural network has been used to classify two types of Malaria. Finally, a graphical user interface has been designed to show result for two types of Malaria. After Complete the designing 95.45% accuracy 90.9%, sensitivity and 100% specificity has been determined.