A Deep Convolutional Neural Network for Detection of Malaria Parasite in Thin Blood Smear Images
M. Antony Robert Raj, Rohan S. Sharma, Deepak Sain · 2021
The Malaria disease parasite detection in laboratories has evolved drastically with advent of Medical Imaging and Artificial Intelligence. This new technologically advance method has proved to be great for medical practitioners because of its independence from external factors such as human intervention and manufacturing defects which may have adverse effects over the results. This paper describes a Deep Learning based image classification approach for the detection of malarial parasite existing in thin blood smear images, using Convolutional Neural Network (CNN) for effective feature extraction and accurate classification. The proposed CNN model can automatically extract intrinsic and discriminative features from images provided. The CNN performs perfectly with image data. This paper annotates a comparison of the observed accuracy of proposed model using three different optimizers on training and validation data. Our custom Deep CNN model can classify the parasitized an uninfected with 93.47 percent accuracy.