Deep Learning Estimation of Small Airways Disease from Inspiratory Chest CT is Associated with FEV Decline in COPD.

Researchers

Journal

Modalities

Models

Abstract

Quantifying functional small airways disease (fSAD) requires additional expiratory computed tomography (CT) scan, limiting clinical applicability. Artificial intelligence (AI) could enable fSAD quantification from chest CT scan at total lung capacity (TLC) alone (fSAD TLC ).To evaluate an AI model for estimating fSAD TLC and study its clinical associations in chronic obstructive pulmonary disease (COPD).We analyzed 2513 participants from the SubPopulations and InteRmediate Outcome Measures in COPD Study (SPIROMICS). Using a subset ( n = 1055), we developed a generative model to produce virtual expiratory CTs for estimating fSAD TLC in the remaining 1458 SPIROMICS participants. We compared fSAD TLC with dual volume, parametric response mapping fSAD PRM . We investigated univariate and multivariable associations of fSAD TLC with FEV 1 , FEV 1 /FVC, six-minute walk distance (6MWD), St. George’s Respiratory Questionnaire (SGRQ), and FEV 1 decline. The results were validated in a subset ( n = 458) from COPDGene study. Multivariable models were adjusted for age, race, sex, BMI, baseline FEV 1 , smoking pack years, smoking status, and percent emphysema.Inspiratory fSAD TLC was highly correlated with fSAD PRM in SPIROMICS (Pearson’s R = 0.895) and COPDGene (R = 0.897) cohorts. In SPIROMICS, fSAD TLC was associated with FEV 1 (L) (adj.β = -0.034, P < 0.001), FEV 1 /FVC (adj.β = -0.008, P < 0.001), SGRQ (adj.β = 0.243, P < 0.001), and FEV 1 decline (mL / year) (adj.β = -1.156, P < 0.001). fSAD TLC was also associated with FEV 1 (L) (adj.β = -0.032, P < 0.001), FEV 1 /FVC (adj.β = -0.007, P < 0.001), SGRQ (adj.β = 0.190, P = 0.02), and FEV 1 decline (mL / year) (adj.β = – 0.866, P = 0.001) in COPDGene. We found fSAD TLC to be more repeatable than fSAD PRM with intraclass correlation of 0.99 (95% CI: 0.98, 0.99) vs. 0.83 (95% CI: 0.76, 0.88).Inspiratory fSAD TLC captures small airways disease as reliably as fSAD PRM and is associated with FEV 1 decline.This work was supported by NHLBI grants R01 HL142625, U01 HL089897 and U01 HL089856, by NIH contract 75N92023D00011, and by a grant from The Roy J. Carver Charitable Trust (19-5154). The COPDGene study ( NCT00608764 ) has also been supported by the COPD Foundation through contributions made to an Industry Advisory Committee that has included AstraZeneca, Bayer Pharmaceuticals, Boehringer-Ingelheim, Genentech, GlaxoSmithKline, Novartis, Pfizer, and Sunovion.