<![CDATA[Malaria is a deadly disease transmitted by the Plasmodium parasite. Detection is performed by trained microscopists who analyze microscopic images of blood smears. This analysis can be done automatically using modern deep learning techniques. The need for skilled labor can be significantly reduced by developing accurate and efficient automated models. In this article, we propose a fully automated convolutional neural network (CNN)-based model for diagnosing malaria from microscopic images of blood smears. Various techniques including knowledge distillation, data augmentation, autoencoder, feature extraction with CNN model to optimize and improve model accuracy and reasoning performance. Our deep learning model can detect malaria parasites from microscopic images with 95% accuracy requiring more than 27,600 images. This shows that the mode is able to provide more accurate predictions compared to malaria disease detection models using other algorithms such as in previous studies with an accuracy of 90%. By using CNN algorithm, this article can contribute novelty in the development of effective malaria detection methods for malaria disease.]]>
Copyrights © 2025
