Malarial Parasite Identification Using Convolution Neural Network
S. Kavitha · Bioscience Biotechnology Research Communications · 2020
Malaria -a dreadful and deadly disease caused by a parasite belong to the plasmodium family that commonly infects a female Anopheles mosquito which bite on humans.With the symptoms, the disease can be diagnosed by trained lab technicians who will examine the microscopic blood smear images.Developing an automatic, accurate and efficient model for detecting this disease will reduce the requirement for the trained human resource and it will improve the diagnosis efficiency.Deep learning neural networks can be used to improve the efficiency and the accuracy of the diagnosis.In this paper, we propose a model using Convolutional Neural Network (CNN) for the examination of malaria from the microscopic human red blood smear images.This model will provide a rapid, accurate, low cost outcome.Our model differentiates the infected and uninfected cell images by training the convolutional neural networks.The algorithm involves the methods and architectures of computer vision, image processing operations and deep learning.The proposed CNN model can examine the malarial parasites from microscopic images with an accuracy of 68.38%, in 10000 checkpoint operations.