MALARIA PARASITE DETECTION USING VGG-16

Abstract

Malaria, caused by the Plasmodium parasite and transmitted through mosquito bites, is a significant global health challenge, with African nations accounting for 94% of cases and deaths. In 2019, the disease caused 409,000 deaths, 67% of which were among children under five. While a precise diagnosis is crucial, conventional techniques—like examining stained blood smears under a microscope—require a lot of resources, including trained staff and lab equipment that is sometimes unavailable in low-resource environments. In order to overcome these obstacles, this work uses deep learning methods for automated malaria identification. Prior machine learning techniques, such as random forests and decision trees, have demonstrated respectable accuracy but are not stable or scalable. The suggested approach makes use of convolutional neural networks (CNNs) and the VGG-16 architecture to enhance diagnostic efficiency and accuracy via transfer learning. In settings with limited resources, this automated methodology provides a scalable and effective method for diagnosing malaria from blood smear pictures by reducing dependency on human expertise. This strategy could revolutionize malaria diagnostics by improving speed and accuracy, resolving significant shortcomings of conventional techniques, and aiding international efforts to eradicate this fatal illness.