DEEP LEARNING MECHANISM FOR RECOGNITION OF MALARIAL PARASITE IN THICK BLOOD SMEAR

Computer Science and Engineering

Malaria is among the deadliest illnesses in the world. This is brought on by a female Anopheles mosquito bite that spreads Plasmodium parasites. Manual microscopic inspection and Rapid Diagnostic Test (RDT) are two modern malaria detection methodologies. These methods are susceptible to errors which made by humans. Worldwide death rates from malaria can be decreased with early detection. For accurately detect the malaria parasite, a thick blood smear examination is required. The World Health Organization (WHO) recommends thick smear screening for the primary diagnosis of malaria because it is affordable and extremely sensitive. Therefore, Deep Learning could be a highly helpful tool for identifying illnesses. This method offers a more expedient and affordable method of identifying the malaria parasite in a blood smear. In this paper, CNN-based machine learning model automatically identifies and predicts infected cells. The primary function of the custom convolutional neural network is to discriminate between blood samples that are healthy and those that are infected. Three fully connected layers plus convolutional layers make up the proposed system. A cascade of several convolutional layers with various filters contained in the layers makes up the proposed neural network, producing unusually good accuracy given the resources at hand. After the model has been trained, many blood sample images are fed into it to check the planned system's correctness. Analysis of blood smear samples can also help in the detection of a few other diseases, and the use of deep learning models will benefit the overall humanity.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.