2015 | OriginalPaper | Chapter
Lung Nodule Detection in X-Ray Images: A New Feature Set
Authors : Burçin Buket Oğul, Polat Koşucu, Ahmet Özçam, Sümeyra Demir Kanik
Published in: 6th European Conference of the International Federation for Medical and Biological Engineering
Publisher: Springer International Publishing
Activate our intelligent search to find suitable subject content or patents.
Select sections of text to find matching patents with Artificial Intelligence. powered by
Select sections of text to find additional relevant content using AI-assisted search. powered by
The high fatality rate of lung cancer brings a lot of attention to Computer Aided Detection (CAD) systems for lung nodule detection. A CAD can help a radiologist to reduce the time, and effort in analyzing images and increase the accuracy. In this paper a fully automated CAD system is presented to detect lung nodules from X-ray images. Proposed system segments the lung area, identifies the nodule candidates, extract the features and classifies the candidates as nodule or not. The output of the system is the highlighted nodule candidate areas with the size information. Publicly available JSRT (Japanese Society of Radiological Technology) images are used to validate the system. We achieved %80 sensitivity with an average of 6.4 FPs. The system is tested on a different dataset with 417 nodules and the sensitivity is %76 with 6.7 FPs. Proposed system shows a potential to fully automate nodule detection from lung X-ray images with satisfying accuracy.