Analysis of Recurrent Neural Network and Convolution Neural Network Techniques in Blood Cell Classification

Abstract

Medical imaging is a process of obtaining photographic images of the human body for the purpose of diagnostics and further treatment. White blood cells are the cells that fight any disease that might want to attack the human body; they are the cells that form the immune system of the body. There are five different types of white blood cells responsible for attacking different invaders, like bacteria and parasites, among others. Recent times have seen medical imaging being used for identification of the different types of white blood cells being produced by the body. Machine learning and deep learning, which have revolutionised many important sectors, have been researched in the field of medical imaging as well. Many research works have been carried out in order to classify the different types of white blood cells, like neutrophils, eosinophils, lymphocytes, and monocytes. Proper identification will lead to better diagnosis of blood-related diseases. Many machine leaning and deep learning techniques have been successfully employed and have been experimented with for this purpose. Deep learning techniques, recurrent neural network and convolution neural network, have been employed for better classification. So, this chapter proposes to conduct a detailed study of different deep learning techniques used for classification of blood cell type and to analyse their effectiveness in producing accurate classification results.