Development of automatic classification system for leukocyte images using Random Forest. Issue 11 (28th August 2018)

• Record Type: Journal Article
• Authors: Tomiyama, Shinnosuke; Sakata‐Yanagimoto, Mamiko; Chiba, Shigeru; Aikawa, Naoyuki
• Is Part Of: Electronics and communications in Japan. Volume 101:Issue 11(2018)
• Page Start: 13
• Page End: 19
• Publication Date: 2018-08-28
• Subjects: blood figure analysis of peripheral blood -- Gram‐Schmidt orthonormalization -- Random Forest; Electronics -- Periodicals; Telecommunication -- Periodicals; 621.6805
• DOI: 10.1002/ecj.12113
• Languages: English
• ISSNs: 1942-9533

Abstract

Abstract: Classifying leukocyte and examining their proportions is very important for disease examination and diagnosis. This examination needs the knowledge of experts and a lot of time. Therefore, many automatic leukocyte image classification algorithms have been proposed. There is a method to classify 13 types of blood cells using 1‐v‐1 support vector machine (SVM) in one of them. In the conventional method, leukocyte images are classified with the 26‐dimensional feature vectors. However, the classification accuracy is poor with these feature vectors in granulocytes in this method. In this article, we propose new feature vectors to improve the classification accuracy of blast cells with low classification accuracy among the leukocyte fractions. That is, we add two feature vectors in the proposed method. And we improve the accuracy of the whole by using a Random Forest for the classifier.