Click here to

Session: CT and CBCT: New Technologies, Algorithms, and Emerging Applications II [Return to Session]

Reproducibility of An Automated Lobar Lung Tissue Assignment Technique Using CT Pulmonary Angiography

N Luu, Y ZHAO, S Molloi*, University of California, Irvine, CA


SU-F-201-4 (Sunday, 7/10/2022) 2:00 PM - 3:00 PM [Eastern Time (GMT-4)]

Room 201

Purpose: Computer tomography pulmonary angiography (CTPA) is a first-line diagnostic technique used to assess patients with suspected pulmonary embolism (PE). Lobar segmentation is necessary for regional ventilation and perfusion analysis. This study evaluated the reproducibility of a vessel-specific minimum-cost path (MCP) technique used for lobar segmentation based on CTPA images.

Methods: A total of nine Yorkshire Swine (52.22±3.19 kg) were used in this study with 18 independent CT pulmonary angiography acquisitions using a 320-Slice CT scanner after intravenous contrast injection (370 mg/mL iodine, 0.5 mL/kg, 5mL/s) followed by a 0.5 mL/kg saline chaser. A helical scan mode was used (100kV, 200 mA, 24 cm Z-coverage and 2.5s scan time). Following image acquisition, lung tissue segmentation and pulmonary arterial tree centerline extraction was performed. The pulmonary arterial tree was then divided into six lobar subtrees for lobar assignment. The MCP technique was used to assign lobar territories by assigning every voxel of the lung tissue to the nearest arterial tree segment. The reproducibility of the MCP technique was evaluated by quantitatively comparing the MCP-derived lobar territories between two CT acquisitions by tissue mass correspondence using linear regression, root mean square error (RMSE), root mean square deviation (RMSD) and paired sample t-test (p-value).

Results: The calculated whole lung mass measurements (469.5±61.8g) from the first (MLUNG1) and second (MLUNG2) CT acquisitions were correlated by MLUNG1 = 0.98MLUNG2 + 8.10g (r=0.98) with a p-value of 0.625, RMSE=12.03g and RMSD=11.76g. The lobar mass measurements from the first (MLobe1) and second (MLobe2) CT acquisitions were correlated by MLOBE1 = 0.98MLOBE2 + 1.03 (r=0.99) with a p-value of 0.455, RMSE=4.51g and RMSD=4.39g.

Conclusion: The whole-lung and lobar mass measurements show excellent reproducibility using a vessel-specific assignment technique. This technique can potentially be used for automated lung lobar segmentation, enabling regional ventilation and perfusion analysis for clinical applications.

Funding Support, Disclosures, and Conflict of Interest: Partially funded through a grant from Canon Medical Systems


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email