A Deep Learning-Based Malarial Parasite Detection Using Blood Smear Images for Healthcare Techniques

Abstract

Malaria is a vector-borne infectious disease that spreads through the bites of infected female mosquitoes, namely Anopheles, infected with the Plasmodium parasite. When an infected mosquito bites a person, the parasite increases its count in the affected person's liver and begins to destroy red blood cells. Traditionally, malaria diagnosis involves visually examining blood under a microscope, but this method can vary based on the expertise and experience of the pathologist. Different types of deep learning techniques have been used to detect infected blood cells automatically to improve diagnosis effectively. However, these methods often require expert knowledge to adjust features for detection. The proposed system of tuning the features using deep learning techniques can accurately detect malaria without needing hand-crafted features. This will be tested on a dataset (blood smear images) that can be accessed by the general public from NIH.