Malaria Detection Using Convolutional Neural Networks: A Comparative Study

The potentially fatal parasitic disease known as malaria is spread to people by mosquito bites. For the disease to be effectively treated and controlled, early malaria detection is essential. Convolutional neural networks (CNNs), a type of deep learning algorithm, have recently demonstrated considerable potential in the detection of malaria from blood smear images. In this study, we assess how well four CNN models-GoogLeNet, DenseNet161, MobileNet_v2, and ResNet18-performed at identifying malaria from a dataset that was made available to the public. Our results show that DenseNet161 achieved the highest accuracy on the test set (95.86%), followed by ResNet18 (95.08%), GoogLeNet (94.85%), and MobileNet_ v2 (94.88%). These results imply that deep learning algorithms are capable of detecting malaria and that selecting the right model is essential for achieving high accuracy.