Due to the remarkable progress of the multi-slice CT scanner, 3 dimensional analyzes of human body using CT images is becoming possible and a lot of image processing procedures are developed to support it The classification of lung region is an important part of such analyzes, because it can provide some effective information for detecting and locating the disease such as lung cancers. However, It is difficult to classify the lung region just by detecting fissure regions using thresholding method because the fissures are not so clear in CT image. In this paper, we proposed a method to classify the lung regions by extracting bronchus and recognizing its structure from multi-slice chest CT images.
展开▼