RSNA 2013 

Abstract Archives of the RSNA, 2013


LL-INE3228-THB

Lung Lobar Segmentation Using an Anatomy-based Priority Knowledge in Low-dose Chest CT: Application to COPD Patients

Education Exhibits

Presented on December 5, 2013
Presented as part of LL-INS-THB: Informatics - Thursday Posters and Exhibits (12:45pm - 1:15pm)

Participants

Sang Joon Park, Presenter: Nothing to Disclose
Jin Mo Goo MD, PhD, Abstract Co-Author: Research Grant, Guerbet SA Research Grant, Toshiba Corporation
Jung Im Kim MD, Abstract Co-Author: Nothing to Disclose
Hyun-Ju Lee MD, PhD, Abstract Co-Author: Nothing to Disclose
Chang Hyun Lee MD, PhD, Abstract Co-Author: Nothing to Disclose
Chang Min Park MD, PhD, Abstract Co-Author: Nothing to Disclose
Sang Min Lee MD, Abstract Co-Author: Nothing to Disclose

BACKGROUND

Lung lobar segmentation in CT images is a challenging tasks because of the limitations in image quality from parenchymal diseases and CT image acquisition, especially low-dose CT for clinical routine environment. The purpose our study was to propose and explore an automatic segmentation technique for pulmonary lobes and to validate its performance with COPD cases

EVALUATION

Thirty COPD patients were selected for investigating the performance of the lobar segmentation scheme in this study. The images were obtained with low-dose chest CT (40 mAs at 120 kVp) using soft reconstruction kernel (Sensation 16). A PC-based in-house software was developed for fully automated segmentation of the pulmonary lobes using the following steps: First, segmentation of airways, vessels and lungs were performed. Then we extracted minor and major fissures by using eigenvalues-ratio of the Hessian matrix. To enhance and recover the faithful 3-D fissure plane, our proposed fissure-enhancing filter were applied to the images. After finishing above steps, for careful smoothening of fissure planes, 3-D rolling-ball algorithm in xy and xz coordinate planes was performed, respectively.

DISCUSSION

By using 30 chest CT data sets, two expert thoracic radiologists performed visual scoring with 5 scales (0: failure, 1: poor, 2: fair, 3: satisfactory, 4: excellent). The mean scores of right and left lungs were 3.63 ± 1.54 (90%) and 3.80 ± 1.09 (95%), respectively. Results show that our proposed scheme showed better results in the left lung than in the right lung. This is due to the fact that 3 cases included large incomplete fissures, another 1 case had a tuberculosis and the others showed fibrotic changes adjacent to the fissure planes in the right lung.

CONCLUSION

By using novel lobar segmentation steps comprising decomposition of fissure planes, we could segment the pulmonary lobes up to 95% success rate even if some cases showed difficult situations for identifying the normal fissures. This study can be a vital role as a preprocessing step for regional analysis including lobes and pulmonary segments in the lung parenchyma for various lung diseases in the clinical environment.

Cite This Abstract

Park, S, Goo, J, Kim, J, Lee, H, Lee, C, Park, C, Lee, S, Lung Lobar Segmentation Using an Anatomy-based Priority Knowledge in Low-dose Chest CT: Application to COPD Patients.  Radiological Society of North America 2013 Scientific Assembly and Annual Meeting, December 1 - December 6, 2013 ,Chicago IL. http://archive.rsna.org/2013/13026617.html