Author: Shiho Amster, Ryan Andosca, Igor Barjaktarevic, Michael Vincent Lauria, Daniel A. Low, Dylan P. O'Connell, Ann Raldow, Brad Stiehl 👨🔬
Affiliation: Department of Pulmonology, University of California Los Angeles, Department of Radiation Oncology, University of California, Los Angeles, Cedars-Sinai Medical Center 🌍
Purpose: Image-based biomarkers could be useful for disease diagnosis and prognosis, especially in heterogeneous lung diseases like COPD. The purpose of this study is to assess the feasibility of using CT-based ventilation as a biomarker for COPD severity.
Methods: We scanned 23 radiotherapy patients with varying COPD status with the 5DCT protocol, consisting of 25 fast-helical free-breathing CT scans with simultaneous monitoring of the breathing waveform. We represented the ventilation as the Jacobian of the deformable vector field between reconstructed end-exhalation and end-inhalation images from the 5DCT model. We calculated the mean ventilation in each lobe across patients within each COPD severity group to demonstrate the relationship between ventilation and COPD. We devised a Ventilation Heterogeneity Index (VHI) to assess the relationship between interlobar heterogeneity and disease severity, where 0 indicates uniform ventilation and higher values reflect a single lobe differing significantly from others. Finally, we crafted a biomarker by relating the percentage of the lungs with voxel-wise Jacobian values below a set of thresholds to COPD severity.
Results: Mean ventilation values decreased with increasing COPD severity across most lobes. The mean VHI values are 0.503 for patients without COPD, 0.658 for mild COPD, 0.631 for moderate-to-severe COPD, 0.643 for all COPD patients, and 0.642 across all patients. All thresholds above 1 show correspondence to disease, but a threshold above 1.1 best differentiated the COPD severity groups. The average percentage of the lungs with expansion values below 1.1 was 31.3% in patients without COPD, 43.3% in mild COPD, and 56.7% in moderate-to-severe COPD.
Conclusion: This study’s results demonstrate the feasibility and value of developing CT-based ventilation into a biomarker for lung disease. Future work will expand the patient cohort and investigate other potential biomarkers derived from ventilation distributions.