Leukocyte (White Blood Cell) Classification with a Multi-stage Support Vector Machine

Department of Electronic Technology and Biomedical Engineering, Hanoi University of Science and Technology, Vietnam

^*Corresponding Author

Duong Trong Luong

Department of Electronic Technology and Biomedical Engineering

Hanoi University of Science and Technology

Vietnam

Email: luong.duongtrong@hust.edu.vn

Abstract

In this paper we present an automatic method for leukocyte classification using support vector machines in multiple stages that results in higher accuracy and less overfitting than other automatic methods. White blood cells (WBC), also called leukocytes, play an important role in the immune system to help protect the body from virus and bacterial diseases. Leukocyte tests can help diagnose a number of diseases related to blood disorders such as acute leukemia, chronic myeloid leukemia. Manual blood cell classification is commonly used in clinics and hospitals, automated cell classification systems may assist clinicians to increase efficiency and accuracy of diagnosis. In recent years, along with the development of artificial intelligence, there have been many studies using automatic methods to classify and count leukocytes. In this paper, we propose a Multi-stage support vector machine method (Multistage SVM) to classify leukocytes into 4 classes with 93% accuracy and less overfitting than other automatic methods.

Keywords: Leukocyte, classification, Support vector machine, Multi-Stage, Blood disorder

Download the full article (PDF)