Malaria Detection Using Contour Detection And Random Forest Classifier

Abstract

The application of machine learning is constantly increasing, especially in the field of automated disease diagnosis and prediction such as malaria which is a serious global health problem, and rapid, accurate diagnosis is required to control the disease. Malaria remains a major burden on global health, with roughly 200 million cases worldwide and more than 400,000 deaths per year. But machine learning can increase the chances of survival by early prognosis and diagnosis if it diagnosed properly and accurately.To improve diagnosis, image analysis software and machine learning methods have been used to quantify parasitemia in microscopic blood slides.In this paper we propose an automated malaria detection approach based on the image analysis using contour detection to detect the contours of the infected cell area which will be five while uninfected cell will have only one and random forest classifier algorithm to build the model which will predict the malaria and provide the report. For this purpose,National Institute of Health,USA malaria dataset is used over which classification rates like precision, recall and F1 score are generated.