A Novel Approach Towards Detecting Malaria Parasite in Thin Blood Smears Using a Sequential Convolutional Neural Network (CNN) Model

Inaccurate or delayed detection of malaria leads to increased casualties. In our proposed work, a novel approach was designed based on a Convolutional Neural Network (CNN) model using a sequential model. A collection of thin blood smears, both infected and uninfected with the malarial parasite, was used as the input, training, and testing of our proposed model. Our proposed approach resulted in an accuracy of 94.6% and a precision of 97.5%, which are significantly higher than those of the existing approaches. This accuracy can be further fine-tuned and increased by increasing the testing and training pools. The goal is to automate the entire malaria detection mechanism to ensure the highest number of true positive results, while minimizing the possibility of human errors or false results. The goal of this study is to achieve optimum accuracy while reducing the amount of time spent on testing.