Automatization of Microscopy Malaria Diagnosis Using Computer Vision and Random Forest Method
Dalibor Đumić, Dino Кеčo, Zerina Mašetić · IFAC-PapersOnLine · 2022
Malaria is an infectious disease caused by Plasmodium species parasites transferred by infected females of Anopheles mosquitoes. Around half of the world population is exposed to some risk of malaria, with the highest incidence in sub-Saharan Africa. High mortality rates and the spread of this disease require a fast and efficient mode of diagnosis. This paper presents a model for automatization of microscopy analysis of malaria suspect blood smears and diagnosis using Random Forest method. The results are evaluated through accuracy, recall and precision metrics, with achieved performance of 90% for all three metrics. This shows the potential of machine learning methods inclusion in the malaria detection process for faster and less costly diagnosis.