A Semi-automatic Extraction Algorithm of Lung Lobar Fissures from HRCT Images Using Ridgelet

Abstract

The pulmonary fissures are the boundaries between the lobes in the lungs. They are useful for the analysis of pulmonary conformation and the diagnosis of lung disease on a lobar level. This paper introduces a technique for the semi-automatic extraction of lung lobar major fissures on HRCT images. First we get the direction and approximate bound of the fissure using Ridgelet transform, and then use converging area method to locate the fissure area. After that, a method based on gradient operator is used to extract the fissure. At last, we can get a continuous fissure using polynomial fit. We applied the proposed algorithm to 200 HRCT images. The average distance between the results of this algorithm and the manually delineated fissures is less than 1.5 mm, within the range of manual extraction variations. The extracted fissures will be aided to diagnose lung cancer and to assess pulmonary emphysema quantified on lobar level automatically.

DOI: 10.1109/BMEI.2008.94

7 Figures and Tables

Cite this paper

@article{Zhang2008ASE, title={A Semi-automatic Extraction Algorithm of Lung Lobar Fissures from HRCT Images Using Ridgelet}, author={Guodong Zhang and Xin Zhang and Hong Zhao and Peiyu Yan}, journal={2008 International Conference on BioMedical Engineering and Informatics}, year={2008}, volume={1}, pages={850-854} }