Blood Cell Counting and Malaria Pathogen Detection using Convolutional Neural Network

The Whole Blood Cell Count is a crucial component of determining a human’s overall health and detecting different abnormalities such as anemia, infection, and leukemia. During this test, the quantity of different blood cells found in an individual is evaluated. Automating this approach would boost diagnosis effectiveness while lowering treatment costs in general. The objective of the paper is to utilize a convolutional neural network to carry out a comprehensive blood cell count through the examination of blood smear images. The model is additionally trained to identify the presence of malarial pathogens in the blood. According to the examinations conducted, the system performs with an average mean accuracy of over 0.95 when compared to the ground truth. Moreover, the model identifies the images that contain malarial parasites as infected with a 100% accuracy rate. Additionally, the software is adapted for use on an inexpensive microcomputer to facilitate fast prototyping.