Automatic Lung Segmentation in Computed Tomography Images Using Active Shape Model
Faridoddin Shariaty, Mahdi Orooji, Mojtaba Mousavi, Maksim Baranov, Elena Velichko
Abstract
Lung segmentation in Computed Tomography (CT) images plays a vital role in the diagnosis, detection and three-dimensional visualization of lung nodules. In addition, the stability, accuracy and efficiency of lung segmentation in CT images have a significant impact on the performance of Computer-Aided Detection (CAD) systems. Lung segmentation is usually the first step in lung CT images analysis. In this paper, a fully automated algorithm for recognition and segmentation the lung in 3D X-ray images using the Active Shape Model (ASM) is presented. Proposed algorithms not only split the left and right lungs automatically, but also include the juxta-pleural nodules as a result of segmentation. This method is based on the ASM algorithm, which automatically detects nodules attached to the lung wall. This algorithm applied to 7 CT images of the lungs that include juxta-pleural nodules and calculate the division dice of segmentation.