Deep Learning for Malaria Diagnosis: Leveraging Convolutional Neural Networks for Accurate Parasite Detection

Abstract

malaria is one of the most severe diseases worldwide. However, the current diagnostic method that involves examining blood smears under a microscope is unreliable and heavily relies on the examiner's expertise. Recent attempts to use deep-learning algorithms for malaria diagnosis have not produced satisfactory results. But, a new CNN-based machine learning model has been proposed in a research paper that can automatically detect and predict infected cells in thin blood smears with 94.63% accuracy. This model accurately accentuates the region of interest for the stained parasite in the images, which increases its reliability, transparency, and comprehensibility, making it suitable for deployment in healthcare settings.