Characteristics determination of infected erithrocytes by plasmodium falciparum as a diagnostic of malaria, based on microscopic images
Izzati Muhimmah, Nani Harniawati, Novyan Lusiyana · 2017
A diagnosis of a stadium of Plasmodium can help physicians in managing patients with malaria. The stadium of Plasmodium is divided into three stadiums, namely trophozoite, schizont, and gametocyte. Staging is based on the morphological characteristics of each stadium, including shape, size, texture, and color. We selected 15 features using a correlation-based feature selection method with a “best first” search algorithm. This system is created to identify the detected object, while the stadium will be determined based on the morphological characteristics of each object. Then, the application will provide information on how many each stadium can successfully identify. This diagnostic tool for malaria based on microscopic images, is intended to identify the characteristics of the stadium of Plasmodium in malaria and classified into trophozoit, schizont or gametocyte. The method used to classify the stadium is a support vector machine (SVM), whose strategy is “one versus all” (OVA), which yields an accuracy of 83.252%. The agreements between the system and the doctor's rating was 74.28% using the single decision threshold.