Classification of Plasmodium Skizon and Gametocytes Malaria Images Using Deep Learning

Abstract

Identification analysis of the malaria parasite cell infection, there is a possibility of human error factor done by paramedics because of the number of samples that must be analyzed. This case is because the human eye tends to be tired while working continuously, which can lead to misclassification and treatment that is not right. Therefore, it takes a computer-based system that facilitates medical expert or laboratory technician in identifying two types of parasite cells namely Plasmodium skizon and Plasmodium gametocytes to reduce instances of human error. This research will be conducted on computer-based identification by processing the image type of plasmodium malariae consists of two types, namely Plasmodium skizon and Plasmodium gametocytes levels using convolutional neural network with VGG-16 pre-trained model using 13 layers and 2 dense layers. This study applied 5-fold cross validation for datasets and the datasets are tested using 4 level epoch nodes. The results showed the success of the classification results which have highest training accuracy 90% as well as the results of the highest testing accuracy 100%• It showed the classification using CNN VGG-16 pre-trained model successfully classified the malaria type images.