Deep-learning methods for the classification of normal and pathological blood cells and bone-marrow cells: A comprehensive review

Our paper comprises a state-of-art study of the Deep-Learning methods for medical imaging and more precisely classification of blood cells based on microscopic images. The Blood composition is considered complex and diverse; however, it remains a pertinent factor in diagnosing a patient’s health. Detecting all blood disorders without Artificial Intelligence necessities years of training and experience. As it takes a lot of time to analyse blood samples, making it harder for modern laboratories to meet high demand. In this context, numerous researches were run to implement AI systems that may assist pathologists and biologists in blood analysis. To date, these provided solutions didn’t meet all the needs of researchers regarding the classification of normal and pathological blood cells. Firstly, these proposed solutions only treat and target specific issues, and each solution has been integrated using a microscope that is dedicated entirely to that specific task. Therefore, a more flexible and general classification system must be implemented. Secondly, smart microscopes are expensive which makes them hard to purchase. In this paper, we cover the proposed methods and models related our context, as well as the data sets that best fit it using Artificial Intelligence and Deep-Learning methods.