Malaria Detection Using Multiple Deep Learning Approaches

With about 200 million global instances and over 400,000 fatalities a year, malaria continues an enormous strain on global health. Modern information technology plays a major part in many attempts to combat the disease, along with biomedical research and political efforts. In specific, insufficient malaria diagnosis was one of the obstacles to a promising mortality decrease. The paper offers an outline of these methods and explores present advancement in the field of microscopic malaria detection and we have ventured into utilization of deep learning for detection of Malaria Parasite. Deep Learning over the years has proven to be much faster and much more accurate as it automates feature extraction of the dataset. In this research paper, we investigated various models of Deep Learning and monitored which of these models provided a better accuracy and faster resolution than previously used deep learning models. Our results show that Resnet 50 model gave the highest accuracy of 0.975504.