Title: Lung Segmentation of X-Ray Thorax Image Using Geometric Active Contour and Enhancement Spatial Domain Filtering

Year of Publication: Apr - 2014
Page Numbers: 368-373
Authors: Mokhamad Amin Hariyadi
Conference Name: The International Conference on Computing Technology and Information Management (ICCTIM2014)
- United Arab Emirates

Abstract:


Chest cavity is an important part of the body structure because there are vital organs in it, namely the lungs. The lungs are the organs that come into contact with the outside of the body, where the disturbances in this organ could have implications for other vital organs, therefore it is necessary for clinical examination and x-ray examination. So far, the observations of lungs from the x-ray image of the thorax have been done manually by a physician or radiologist, the results obtained are generally subjective. Therefore, it is needed to have medical aids that can provide a better analysis and used in various things, such as mass screening thorax. In this research, an application for segmenting lungs from the thorax x-ray image is developed using geometric active contour and enhancement spatial domain fitering. This study used 45 X-ray images of the chest cavity (thorax). Results of performance testing in the geometric active contour segmentation, it is obtained: accuracy of 95.2%, sensitivity of 70.4% and specificity of 99.8% for the average right and the left lungs.