Detection Method of Chest X-ray Lung Nodules Based on Improved YOLOV3 Model
Science Discovery
Volume 8, Issue 1, February 2020, Pages: 18-23
Received: Apr. 11, 2020; Published: May 19, 2020
Views 131      Downloads 22
Author
Yan Sitao, School of Software Engineering, Beijing University of Posts and Telecommunications, Beijing, China
Article Tools
Follow on us
Abstract
Lung cancer is a kind of cancer with high lethality in the world, and the detection of lung nodules becomes particularly important in the early clinical manifestations of lung cancer. However, due to the characteristics of the small size of the lung nodules on the front of the chest and the obstruction of the ribs, it is more difficult to manually detect the lung nodules. At the same time, because of the explosive growth of lung X-ray images and diagnostic reports, the The application of deep learning technology to the identification of lung nodules has become an inevitable. In order to be able to detect lung nodules in real time, this paper is based on the YOLOV3 algorithm. In view of the characteristics of lung nodules imaging, such as small size and rib occlusion, a method that can be used Lung nodule detection algorithm (nodule-YOLOV3), the pre-processed lung nodule X-ray image is input to the nodule-YOLOV3 network to obtain the prediction results of lung nodules. The experimental results show that: nodule -YOLOV3 detection accuracy is 61%, compared with YOLOV3 target detection algorithm, the accuracy of nodule-YOLOV3 algorithm is improved by 3%.
Keywords
Lung Nodule, X-ray Images, Deep Learning, YOLOV3
To cite this article
Yan Sitao, Detection Method of Chest X-ray Lung Nodules Based on Improved YOLOV3 Model, Science Discovery. Vol. 8, No. 1, 2020, pp. 18-23. doi: 10.11648/j.sd.20200801.15
Copyright
Copyright © 2020 Authors retain the copyright of this article.
This article is an open access article distributed under the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/) which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ADDRESS
Science Publishing Group
1 Rockefeller Plaza,
10th and 11th Floors,
New York, NY 10020
U.S.A.
Tel: (001)347-983-5186