Abstract

Info

Author(s):: SUN Kai1; 2; YAO Xufeng2; MA Fengling1; 2; ZHAO Wenshuo1; 2; HUANG Gang2; 1. School of Medical Instrument and Food Engineering, University of Shanghai for Science and Technology, Shanghai 200082, China; 2. College of Medical Imaging, Shanghai University of Medicine & Health Sciences, Shanghai 200120, China

Keywords:: Keywords: machine learning; image processing; classification; blood cell

Abstract:: The study of machine learning (ML) for blood cell classification has aroused the interests of many researchers. In this paper, we summarized the recent development of ML algorithms for blood cell classification. The reviewed ML algorithms mainly consisted of data acquisition, image prepossessing, image segmentation, feature extraction and classification. Derived from traditional ML algorithms, the deep learning (DP) algorithms for blood cell classification have demonstrated strong prospects for presenting the advantages of high accuracy and more reliability. Till now, the topics of DP methods focuses on the aspects of extraction of artificial feature, design of learning networks, etc. This would aims to improve the accuracy of classification and generalization of DP models. However, ML classification of blood cells still have some challenges for clinical applications.