Related Experiment Video
Updated: Jul 2, 2026
Lung CT Segmentation to Identify Consolidations and Ground Glass Areas for Quantitative Assesment of SARS-CoV Pneumonia
Published on: December 19, 2020
Automated CT Image Processing for the Diagnosis, Prediction, and Differentiation of Phenotypes in Chronic Lung
Stefan Kuhnert1,2, Nermin Halim1, Janine Sommerlad1
1Department of Internal Medicine II, University Hospital Giessen and Marburg, Justus-Liebig-University Giessen, Giessen, Germany.
Background:
Chronic lung allograft dysfunction (CLAD) after lung transplantation is a common complication with a poor prognosis. We assessed the utility of quantitative computed tomography (CT) for the diagnosis, prediction, and discrimination of CLAD phenotypes.
Methods:
We retrospectively analyzed routine inspiratory and expiratory CT scans from 78 patients at different time points after lung transplantation. Mean lung density (MLD), parametric response mapping (PRM), percentage of air trapping, and airway wall morphology parameters were calculated using the image processing software YACTA. Diagnostic and predictive utility was determined by receiver operating characteristic analysis and Pearson correlation.
Results:
Markers of air trapping showed promise for the diagnosis and prediction of bronchiolitis obliterans syndrome (BOS); for example, expiratory MLD showed areas under the curve (AUCs) of 0.905 for diagnosis and 0.729 for 1-year prediction. For diagnosis of CLAD with mixed phenotype, peripheral measurements (e.g., PRM of peripheral functional small airway disease: AUC 0.893) were most suitable. Markers of airway thickening (e.g., expiratory wall thickness at an inner perimeter of 10 mm: AUC 0.767) gave good diagnostic values for the undefined phenotype. CT biomarkers differed significantly among CLAD phenotypes.
Conclusions:
Different CT biomarkers are suitable for the diagnosis of CLAD phenotypes, prediction of BOS, and differentiation of CLAD phenotypes.
