Classification of Lungs Nodules using Hybrid Features and Neural Network
Lungs nodule detection and classification is a very crucial step for computer aided diagnosis (CAD) systems. In this paper, the authors have proposed a CAD system that consists of multiple phases. In the first phase, Lungs segmentation has been performed. After that, region of interest ROl that cont...
Saved in:
Published in | International Information Institute (Tokyo). Information Vol. 17; no. 5; p. 1771 |
---|---|
Main Authors | , , |
Format | Journal Article |
Language | English |
Published |
Koganei
International Information Institute
01.05.2014
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | Lungs nodule detection and classification is a very crucial step for computer aided diagnosis (CAD) systems. In this paper, the authors have proposed a CAD system that consists of multiple phases. In the first phase, Lungs segmentation has been performed. After that, region of interest ROl that contains nodule has been extracted. Different types of features have been extracted for nodule classification. Artificial Neural Network has been used for classification. The technique was tested on the dataset Lungs Image Database Consortium LIDC and images taken from Aga Khan Medical University. The results confirm the validity of the technique, as well as its enhanced performance. |
---|---|
Bibliography: | ObjectType-Article-2 SourceType-Scholarly Journals-1 ObjectType-Feature-1 content type line 23 |
ISSN: | 1343-4500 1344-8994 |