CALM-QE - Pilot Study on AI-Based Quantitative CT Thorax Analysis in Stable and Exacerbation Phases of COPD.

<bold>Background:</bold> Airway Wall Thickness (10 mm Internal Perimeter Pi10) and Low-Attenuation Area (LAA) Volume (%) have been proposed as biomarkers in diagnosis and predicting COPD prognosis. <bold>Method:</bold> CALM-QE, a multicenter project involves 12 German university hospitals and affiliated outpatient clinics (Sponsor Deutsches Zentrum für Luft- und Raumfahrt e.V.), commissioned by Bundesministerium für Bildung und Forschung, GRANT NUMBER 01ZZ2318A. A total of 19,000 chest CT scans from eight centers will be analyzed. A pilot study randomly selected 180 CT scans, primarily exacerbation cases, out of 1,806 scans at the University Hospital of Freiburg (2017–2024), focusing on Pi10 (mm) and LAA (%). <bold>Results:</bold> The AI software Coreline AVIEW generated 32,733,484 parameters of 180 CT scans: the mean Pi10 was 4.938 mm, median 4.82 mm (SD: ±0.95). Pi10 values >2.6 mm have been linked to exacerbations and mortality, proposed cut-off 2.6 mm. The mean LAA was 17.811% (SE: ±1.0755), median 9.093%. Literature values vary, one study identifies a LAA >5.6% in ROC analysis, indicating COPD progression. <bold>Conclusion:</bold> This pilot study shows that elevated Pi10 and LAA values correlate with COPD exacerbations and hightlights the potential of AI-based quantitative CT analysis in assessing COPD severity and exacerbations. Further analysis within the CALM-QE study will provide deeper insights into these biomarkers, enhancing risk assessment and clinical decision-making in COPD management.