Malaria Detection by The Use of Machine Learning and Deep Learning

Malaria is currently the worst illness on Earth, and dealing with it is a major and difficult task for the health department. A lab or skilled technician visually inspects human blood smears for parasite-infected red blood cells, which is the usual way of diagnosing malaria. This procedure is tedious, time-consuming and requires a skilled pathologist to perform the examination. Blood smears for malaria have previously been diagnosed using deep learning algorithms. But thus far, practical performance has fallen short. This study proposes an automated framework driven by deep learning, where Infected cells in thin blood smears on conventional microscope slides can be automatically classified into two classes. Using 27,000 images of individual cells, to train a convolutional neural network with a 5-fold cross-validation layer to find the cell's parameter. Three different CNN model types—Basic CNN, VGG-19 Frozen CNN, and VGG-19 Fine-Tuned CNN—are evaluated according to their accuracy in order to determine which is the most accurate. The model with the highest accuracy rate is then obtained by comparing the three models’ accuracy.