Classification of plasmodium falciparum based on textural and morphological features

Abstract

Malaria is a disease caused by plasmodium parasites transmitted through the bites of female anopheles-mosquito that infect the human red blood cell (RBC). The standard malaria diagnosis is based on manual examination of a thick and thin blood smear, which heavily depends on the microscopist experience. This study proposed a system that can identify the life stages of plasmodium falciparum in human RBC. The image preprocessing process was done by illumination correction using gray world assumption, contrast enhancement using shadow correction, extraction of saturation component, and noise filtering. The segmentation process was applied using Otsuthresholding and morphological operation. The test results showed that the use of artificial neural network (ANN) using a combination of texture and morphological features gives better results when compared to the use of only texture or morphology features. The results showed that the proposed feature achieved an accuracy of 82.67%, a sensitivity of 82.18%, and a specificity of 94.17%, thus improving decision-making for malaria diagnosis.