Detection of Malaria Infection Using Convolutional Neural Networks

A parasite that causes the potentially fatal disease malaria is spread to humans by female Anopheles mosquitoes carrying this parasite. Despite global efforts to eradicate the disease, malaria is still a major public health problem, particularly in underdeveloped nations. The timely and accurate detection of malaria infection is necessary for effective treatment and for reducing the morbidity and mortality associated with the disease. However, identifying malaria-infected cells can be challenging, especially in areas with limited access to trained healthcare professionals. Recent advances in deep learning and computer vision have shown promising results in identifying several medical images. This paper proposes a new CNN model to identify whether cells are parasitized or uninfected with malaria to help health personnel save the lives of infected people. In various ways, the proposed CNN architecture is compared to other pre-trained models, VGG-19, ResNet50, DenseNet121, and Inception V3. The malaria cell dataset of 27,558 images is used for model evaluations. The findings show that the proposed model performs the best, with an accuracy score of 97%.