Consistency Between Two Independent Methods of Lobe-Wise Ventilation Calculation From Free-Breathing CT
M Lauria*, B Stiehl, D O'Connell, A Santhanam, L Naumann, P Boyle, I Barjaktarevic, D Low, UCLA, Los Angeles, CA
Presentations
PO-GePV-M-179 (Sunday, 7/10/2022) [Eastern Time (GMT-4)]
ePoster Forums
Purpose: The assessment of ventilation as a measure of lung function is essential in function sparing treatment planning for radiation therapy and pulmonary surgeries. However, calculating ventilation accurately is traditionally difficult. The purpose of this study is to investigate two novel approaches to calculating ventilation on free-breathing CT images.
Methods: For 34 lung cancer patients, we acquired 25 free-breathing CT scans and simultaneously monitored the breathing waveform with a respiratory bellows surrogate. Lobe segmentations were performed with the Pulmonary Toolkit in Matlab. For the first ventilation approach, we constructed 5DCT breathing motion models and used them to obtain deformation vector fields between end-exhalation and end-inhalation amplitudes. The mean Jacobian of this vector field within each lobe represented its ventilation. For the second approach, we calculated the Jacobians of 24 CT registrations to a common reference scan. We averaged the Jacobians over slabs of 3 slices. We calculated a linear fit of the slab Jacobians to the amplitudes of the middle slice across the 24 registrations with the amplitudes normalized to the breathing range. The volume-weighted averages of the slopes within each lobe represented its ventilation using this method. Testing the consistency between the two proposed techniques was a way to study the effectiveness of these methods without having a ground truth.
Results: The mean absolute ventilation difference between the two techniques was 0.012 ± 0.038. The ventilation discrepancies were independent of COPD severity.
Conclusion: We determined that the two approaches resulted in consistent ventilation assessment results at the lobular level. Lobe-wise analysis provides a step towards regional ventilation analysis that could be useful in guiding interventions or function sparing treatment planning for radiotherapy. Further studies will aim at correlating the lobe-wise ventilation to disease characteristics.
Funding Support, Disclosures, and Conflict of Interest: This work is supported by the Tobacco Related Disease Research Program 27IR 0056, NIH R56 1R56HL139767 01A1, Ken and Wendy Ruby Foundation, and the UCLA Department of Radiation Oncology.
Keywords
Ventilation/perfusion, Lung, Functional Imaging
Taxonomy
IM/TH- Image Analysis (Single Modality or Multi-Modality): Quantitative imaging
Contact Email