Genomic Insights into the Classification and Identification of Plasmodium Parasites

Malaria is a contagious disease caused by plasmodium parasites transmitted to humans through infected female anopheles mosquito bites. The parasites invade Red Blood Cells (RBCs), impairing oxygen transport and potentially leading to respiratory distress and serious complications. Accurate and timely detection of plasmodium-infected cells is crucial for effective treatment and disease management. This study explores the application of Feed Forward Neural Network (FFNN) for the classification of plasmodium parasite-infected cells versus uninfected cells. Our Work Leverages a dataset comprising 27,558 colored images. We preprocess the data using Bilateral filtering, Histogram equalization, and Sobel filters to enhance model robustness. Our FFNN model, characterized by multiple hidden layers, and activation functions, is trained to discern subtle differences between infected and uninfected cells. Gaussian filter is used to highlight and detect the parasitized regions. The model’s performance is evaluated using classification rate, precision, Log Loss etc. Initial results demonstrate that our model classifies the parasitized and uninfected cells with a classification rate of 97.5%, precision of 97.62, and log loss of 1.857.