This paper presents a method for extracting bronchus regions from 3D chest CT images. Several groups have proposed methods for extracting bronchus regions by using region growing. However, these methods do not work well in the case of tumors dividing the bronchial tree into several parts. This study proposes a method that can extract the bronchus in such cases using local intensity structure analysis, CT value distribution and shape features. First, the proposed method extracts cylindrical structures by using the hessian matrix to analyze local intensity structures. Second, we extract tube structures by using a modified RRF (Radial Reach Filter) to analyze CT value distributions and shape features. Finally, we extract the bronchus by combining the result of RRF and hessian analysis. We applied the proposed method to four 3D chest CT datasets and show that this method can extract well in the case of tumors dividing the bronchial tree.     «

This paper presents a method for extracting bronchus regions from 3D chest CT images. Several groups have proposed methods for extracting bronchus regions by using region growing. However, these methods do not work well in the case of tumors dividing the bronchial tree into several parts. This study proposes a method that can extract the bronchus in such cases using local intensity structure analysis, CT value distribution and shape features. First, the proposed method extracts cylindrical struc...     »