White Blood Cells Classification with Deep Convolutional Neural Networks

Abstract

The necessary step in the diagnosis of leukemia by the attending physician is to classify the white blood cells in the bone marrow, which requires the attending physician to have a wealth of clinical experience. Now the deep learning is very suitable for the study of image recognition classification, and the effect is not good enough to directly use some famous convolution neural network (CNN) models, such as AlexNet model, GoogleNet model, and VGGFace model. In this paper, we construct a new CNN model called WBCNet model that can fully extract features of the microscopic white blood cell image by combining batch normalization algorithm, residual convolution architecture, and improved activation function. WBCNet model has 33 layers of network architecture, whose speed has greatly been improved compared with the traditional CNN model in training period, and it can quickly identify the category of white blood cell images. The accuracy rate is 77.65% for Top-1 and 98.65% for Top-5 on the training set, while 83% for Top-1 on the test set. This study can help doctors diagnose leukemia, and reduce misdiagnosis rate.

Deep learning approach to peripheral leukocyte recognition

“…The dual-stage CNN architecture classified images into 10 types of myeloid and erythroid maturation series, and achieved excellent performance. Moreover, based on deep residual learning theory and medical domain knowledge, Qin et al [26, 27] presented a fine-grained leukocyte classification method for microscopic images. This proposed deep residual neural network was tested on microscopic image dataset with 40 leukocyte categories, and achieved desired results.…”

A Survey On Automated Methods Used For WBC Classification

“…WBCNet's rate of convergence was faster, the training data fits better, and the accuracy was higher. Ultimately, the experiment concluded that WBCNet was overall better than the customary models [4].…”

Technologies for Developing Decision Support Systems for the Diagnosis of Blood Disorders Using Convolutional Neural Networks

“…Суть технологии глубокого обучения заключается в том, что путь извлечения признаков не разрабатывается людьми, а изучается на основе данных с использованием процедуры обучения общего назначения. В области глубокого обучения, конволюционные нейронные сети (CNN [24][25][26][27][29][30], которая должна обеспечить наличие во входном изображении кандидатов на объекты по сегментации, а количество объектов не превышает одного за счет обрезки изображения вручную или сложного шага сегментации. Эти классификационно-задачные методы, как правило, направлены на распознавание пяти типов зрелых лейкоцитов, обычно встречающихся в периферической крови, и начинают классификацию с обрезанных формами лейкоцитов, что приводит к неудобствам в реальных приложениях.…”