A Novel Lung Segmentation and Boundary Correction Method

Segmentation of lung region in chest-computed tomography(CT) is often performed as a preprocessing step in lung imaging. This task is complicated especially in present of disease. In this article, we presented a novel lung segmentation algorithm. We combined region growing and thresholding to do rough segmentation, and used a boundary correction algorithm based on convex hull operation and ray casting to do boundary correction. Its novelty lies in the fact that it corrects the lung border in a geometric way and can reliable include juxtapleural nodules and pulmonary vessels while minimizing oversegmentation of adjacent regions such as the abdomen and mediastinum. Experiments showed that our method effectively incorporates pleural nodules and pulmonary vessels into the segmentation results, the segmentation accuracy has improved.