Lung Compliance Imaging: A Novel Imaging Parameter to Evaluate Lung Stiffness

Document Type

Conference Proceeding - Restricted Access

Publication Date


Publication Title

American Journal of Respiratory and Critical Care Medicine


Rationale: Automated lung quantification methods rely on “Machine learning” models based on expert radiologist consensus. However, consensus training risks an inability to discern changes beyond visually identifiable patterns. We computed the “True regional volume changes” based on 4 Dimensional CT derived ventilation (CT-V). Methods: We prospectively compared the regional lung ventilation (Delta V) between 9 patients with Idiopathic pulmonary fibrosis (IPF) and 6 patients with lung nodules under two different pressure settings: 5cm and 10 cm H20 CPAP during free breathing. Delta volume changes between inspiration and expiration were computed at a voxel level from the corresponding CT-V images. CT-V was computed with the “Integrated Jacobian Formulation” (IJF), which estimates breathing induced volume changes with CT image processing methods (lung segmentation, deformable image registration). The right and left lung were divided into three equal zones based on the lung mask. We denoted an area as “Stiff” if the Delta one minus the V 5/10 ratio was positive. Results: IPF patients were older (72 vs. 66, p 0.05) and had lower forced vital capacity (FVC) (76.2 vs. 97, p 0.01) compared to patients with lung nodules (Table-1). Patients with IPF had increased right lung stiffness compared to the lung nodule group (78% vs. 33%), but this did not achieve statistical significance. The right and left upper lung zones had increased stiffness in IPF patients. FVC has an overall poor correlation with right and left lung compliance scores (Spearman correlation coefficients 0.20 and 0.21). However, within each lung, upper lung zones correlated better to FVC (right 0.48 and left 0.48) compared to lower lung zones (right 0.23 and left 0.08). On multivariate regression analysis for predictive factors of total lung compliance, only age (p 0.004) and RUL stiffness (p 0.008) were significant after adjusting for all other variables. Conclusion: Our study shows for the first time differential regional lung stiffness in patients with IPF and lung nodule. Contrary to popular belief, that lower lung zones are stiffer in IPF patients, our results indicate upper lung zones with minimal scarring to be less compliant. This is a pilot study showing “proof of concept”, and we anticipate a larger longitudinal study to delineate and accurately predict outcomes based on lung stiffness.




Online Abstracts

First Page


Last Page



American Thoracic Society 2019 International conference_restricted, May 17-22, 2019 - Dallas, TX. Meeting Abstract: A5887