Advertisement

A Convolutional Neural Network Approach to Quantify Lung Disease Progression in Patients with Fibrotic Hypersensitivity Pneumonitis (HP)

Published:November 16, 2021DOI:https://doi.org/10.1016/j.acra.2021.10.005
      Rationale and Objectives To evaluate associations between longitudinal changes of quantitative CT parameters and spirometry in patients with fibrotic hypersensitivity pneumonitis (HP).
      Materials and Methods Serial CT images and spirometric data were retrospectively collected in a group of 25 fibrotic HP patients. Quantitative CT analysis included histogram parameters (median, interquartile range, skewness, and kurtosis) and a pretrained convolutional neural network (CNN)-based textural analysis, aimed at quantifying the extent of consolidation (C), fibrosis (F), ground-glass opacity (GGO), low attenuation areas (LAA) and healthy tissue (H).
      Results At baseline, FVC was 61(44-70) %pred. The median follow-up period was 1.4(0.8-3.2) years, with 3(2-4) visits per patient. Over the study, 8 patients (32%) showed a FVC decline of more than 5%, a significant worsening of all histogram parameters (p≤0.015) and an increased extent of fibrosis via CNN (p=0.038). On histogram analysis, decreased skewness and kurtosis were the parameters most strongly associated with worsened FVC (respectively, r2=0.63 and r2=0.54, p<0.001). On CNN classification, increased extent of fibrosis and consolidation were the measures most strongly correlated with FVC decline (r2=0.54 and r2=0.44, p<0.001).
      Conclusion CT histogram and CNN measurements provide sensitive measures of functional changes in fibrotic HP patients over time. Increased fibrosis was associated with FVC decline, providing index of disease progression. CNN may help improve fibrotic HP follow-up, providing a sensitive tool for progressive interstitial changes, which can potentially contribute to clinical decisions for individualizing disease management.

      Key Words

      Abbreviations:

      AIC (Akaike information criterion), ATS (American thoracic society), HP (hypersensitivity pneumonitis), CNN (convolutional neural network), DLCO (Diffusion capacity for carbon monoxide), ERS (European respiratory society), FVC (forced vital capacity), FEV1 (forced expiratory volume in 1 second), C (consolidation), F (fibrosis), GGO (ground-glass opacities), H (normal), LAA (low attenuation areas)
      To read this article in full you will need to make a payment

      Purchase one-time access:

      Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online access
      One-time access price info
      • For academic or personal research use, select 'Academic and Personal'
      • For corporate R&D use, select 'Corporate R&D Professionals'

      Subscribe:

      Subscribe to Academic Radiology
      Already a print subscriber? Claim online access
      Already an online subscriber? Sign in
      Institutional Access: Sign in to ScienceDirect

      References

        • Vourlekis JS
        • Schwarz MI
        • Cherniack RM
        • et al.
        The effect of pulmonary fibrosis on survival in patients with hypersensitivity pneumonitis.
        Am J Med. 2004; 116: 662-668https://doi.org/10.1016/j.amjmed.2003.12.030
        • Churg A
        • Sin DD
        • Everett D
        • et al.
        Pathologic patterns and survival in chronic hypersensitivity pneumonitis.
        Am J Surg Pathol. 2009; 33: 1765-1770https://doi.org/10.1097/PAS.0b013e3181bb2538
        • Raghu G
        • Wilson KC
        • Bargagli E
        • et al.
        Diagnosis of hypersensitivity pneumonitis in adults: an official ATS/JRS/ALAT clinical practice guideline.
        Am J Respir Crit Care Med. 2020; 202: e36-e69https://doi.org/10.1164/rccm.202005-2032ST
        • Raghu G
        • Collard HR
        • Anstrom KJ
        • et al.
        Idiopathic pulmonary fibrosis: clinically meaningful primary endpoints in phase 3 clinical trials.
        Am J Respir Crit Care Med. 2012; 185: 1044-1048https://doi.org/10.1164/rccm.201201-0006PP
        • Fiddler CA
        • Simler N
        • Thillai M
        • et al.
        Use of mycophenolate mofetil and azathioprine for the treatment of chronic hypersensitivity pneumonitis—a single-centre experience.
        Clin Respir J. 2019; 13: 791-794https://doi.org/10.1111/crj.13086
        • Wells AU
        • Flaherty KR
        • Brown KK
        • et al.
        Nintedanib in patients with progressive fibrosing interstitial lung diseases—subgroup analyses by interstitial lung disease diagnosis in the INBUILD trial: a randomised, double-blind, placebo-controlled, parallel-group trial.
        Lancet Respir Med. 2020; 8: 453-460https://doi.org/10.1016/S2213-2600(20)30036-9
        • Procter AJ
        • Jacob J
        Visual vs. computer-based computed tomography analysis for the identification of functional patterns in interstitial lung diseases.
        Curr Opin Pulm Med. 2019; 25: 426-433https://doi.org/10.1097/MCP.0000000000000589
        • Selman M
        • Pardo A
        • King TE
        Hypersensitivity pneumonitis: insights in diagnosis and pathobiology.
        Am J Respir Crit Care Med. 2012; 186: 314-324https://doi.org/10.1164/rccm.201203-0513CI
        • Dias OM
        • Baldi BG
        • Pennati F
        • et al.
        Computed tomography in hypersensitivity pneumonitis: main findings, differential diagnosis and pitfalls.
        Expert Rev Respir Med. 2018; 12: 5-13https://doi.org/10.1080/17476348.2018.1395282
        • Hanak V
        • Golbin JM
        • Hartman TE
        • et al.
        High-resolution CT findings of parenchymal fibrosis correlate with prognosis in hypersensitivity pneumonitis.
        Chest. 2008; https://doi.org/10.1378/chest.07-3005
        • Walsh SLF
        • Calandriello L
        • Sverzellati N
        • et al.
        Interobserver agreement for the ATS/ERS/JRS/ALAT criteria for a UIP pattern on CT.
        Thorax. 2016; 71: 45-51https://doi.org/10.1136/thoraxjnl-2015-207252
        • Best AC
        • Meng J
        • Lynch AM
        • et al.
        Idiopathic pulmonary fibrosis: physiologic tests, quantitative CT indexes, and CT visual scores as predictors of mortality.
        Radiology. 2008; 246: 935-940https://doi.org/10.1148/radiol.2463062200
        • Best AC
        • Lynch AM
        • Bozic CM
        • et al.
        Quantitative CT indexes in idiopathic pulmonary fibrosis: relationship with physiologic impairment.
        Radiology. 2003; https://doi.org/10.1148/radiol.2282020274
        • Uppaluri R
        • Hoffman EA
        • Sonka M
        • et al.
        Computer recognition of regional lung disease patterns.
        Am J Respir Crit Care Med. 1999; https://doi.org/10.1164/ajrccm.160.2.9804094
        • Jacob J
        • Bartholmai BJ
        • Rajagopalan S
        • et al.
        Predicting outcomes in idiopathic pulmonary fibrosis using automated computed tomographic analysis.
        Am J Respir Crit Care Med. 2018; 198: 767-776https://doi.org/10.1164/rccm.201711-2174OC
        • Humphries SM
        • Yagihashi K
        • Huckleberry J
        • et al.
        Idiopathic pulmonary fibrosis: data-driven textural analysis of extent of fibrosis at baseline and 15-month follow-up.
        Radiology. 2017; 285: 270-278https://doi.org/10.1148/radiol.2017161177
        • Kim HJ
        • Brown MS
        • Chong D
        • et al.
        Comparison of the quantitative CT imaging biomarkers of idiopathic pulmonary fibrosis at baseline and early change with an interval of 7 months.
        Acad Radiol. 2015; https://doi.org/10.1016/j.acra.2014.08.004
        • Kim HJ
        • Tashkin DP
        • Clements P
        • et al.
        A computer-aided diagnosis system for quantitative scoring of extent of lung fibrosis in scleroderma patients.
        Clin Exp Rheumatol. 2010; 28: S26-SS3
        • Maldonado F
        • Moua T
        • Rajagopalan S
        • et al.
        Automated quantification of radiological patterns predicts survival in idiopathic pulmonary fibrosis.
        Eur Respir J. 2014; 43: 204-212https://doi.org/10.1183/09031936.00071812
        • Aliboni L
        • Dias OM
        • Pennati F
        • et al.
        Quantitative CT analysis in chronic hypersensitivity pneumonitis: a convolutional neural network approach.
        Acad Radiol. 2020; https://doi.org/10.1016/j.acra.2020.10.009
        • Bermejo-Peláez D
        • SY Ash
        • Washko GR
        • et al.
        Classification of interstitial lung abnormality patterns with an ensemble of deep convolutional.
        Neural Networks. Sci Rep. 2020; 10: 338https://doi.org/10.1038/s41598-019-56989-5
        • Walsh SLF
        • Calandriello L
        • Silva M
        • et al.
        Deep learning for classifying fibrotic lung disease on high-resolution computed tomography: a case-cohort study.
        Lancet Respir Med. 2018; 6: 837-845https://doi.org/10.1016/S2213-2600(18)30286-8
        • Anthimopoulos M
        • Christodoulidis S
        • Ebner L
        • et al.
        Lung pattern classification for interstitial lung diseases using a deep convolutional neural network.
        IEEE Trans Med Imaging. 2016; 35: 1207-1216https://doi.org/10.1109/TMI.2016.2535865
        • De Castro Pereira CA
        • Sato T
        • Rodrigues SC
        New reference values for forced spirometry in white adults in Brazil.
        J Bras Pneumol. 2007; 33: 397-406https://doi.org/10.1590/s1806-37132007000400008
        • Depeursinge A
        • Vargas A
        • Platon A
        • et al.
        Building a reference multimedia database for interstitial lung diseases.
        Comput Med Imaging Graph. 2012; 36: 227-238https://doi.org/10.1016/j.compmedimag.2011.07.003
        • Hansell DM
        • Bankier AA
        • MacMahon H
        • et al.
        Fleischner Society: glossary of terms for thoracic imaging.
        Radiology. 2008; 246: 697-722https://doi.org/10.1148/radiol.2462070712
        • Pennati F
        • Salito C
        • Aliverti A
        Registration of lung CT images acquired in different respiratory ranges with 4DCT and HRCT.
        in: Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society. EMBS, 2015: 2936-2939
        • Jerman T
        • Pernus F
        • Likar B
        • Spiclin Z
        Enhancement of vascular structures in 3D and 2D angiographic Images.
        IEEE Trans Med Imaging. 2016; 35: 2107-2118https://doi.org/10.1109/TMI.2016.2550102
        • Jacob J
        • Bartholmai BJ
        • Egashira R
        • et al.
        Chronic hypersensitivity pneumonitis: identification of key prognostic determinants using automated CT analysis.
        BMC Pulm Med. 2017; 17: 1-12https://doi.org/10.1186/s12890-017-0418-2
      1. Chollet F (2015) Keras

        • Abadi M
        • Barham P
        • Chen J
        • et al.
        TensorFlow: a system for large-scale machine learning.
        in: Proceedings of the 12th USENIX Conference on Operating Systems Design and Implementation. USENIX Association, Berkeley, CA, USA. 2016: 265-283
        • Burnham KP
        • Anderson DR
        Multimodel inference: understanding AIC and BIC in model selection.
        Sociol Methods Res. 2004; 33: 261-304https://doi.org/10.1177/0049124104268644
        • Camiciottoli G
        • Orlandi I
        • Bartolucci M
        • et al.
        Lung CT densitometry in systemic sclerosis: correlation with lung function, exercise testing, and quality of life.
        Chest. 2007; https://doi.org/10.1378/chest.06-1401
        • Yoon RG
        • Seo JB
        • Kim N
        • et al.
        Quantitative assessment of change in regional disease patterns on serial HRCT of fibrotic interstitial pneumonia with texture-based automated quantification system.
        Eur Radiol. 2013; 23: 692-701https://doi.org/10.1007/s00330-012-2634-8
        • Jacob J
        • Bartholmai BJ
        • Rajagopalan S
        • et al.
        Unclassifiable-interstitial lung disease: outcome prediction using CT and functional indices.
        Respir Med. 2017; https://doi.org/10.1016/j.rmed.2017.07.007
        • Salisbury ML
        • Lynch DA
        • Van Beek EJR
        • et al.
        Idiopathic pulmonary fibrosis: the association between the adaptive multiple features method and fibrosis outcomes.
        Am J Respir Crit Care Med. 2017; 195: 921-929https://doi.org/10.1164/rccm.201607-1385OC
        • Jacob J
        • Bartholmai BJ
        • Rajagopalan S
        • et al.
        Serial automated quantitative CT analysis in idiopathic pulmonary fibrosis: functional correlations and comparison with changes in visual CT scores.
        Eur Radiol. 2018; 28: 1318-1327https://doi.org/10.1007/s00330-017-5053-z
        • Lecun Y
        • Bengio Y
        • Hinton G
        Deep learning.
        Nature. 2015; 521: 436-444https://doi.org/10.1038/nature14539
        • Choe J
        • Lee SM
        • Do KH
        • et al.
        Deep learning–based image conversion of CT reconstruction kernels improves radiomics reproducibility for pulmonary nodules or masses.
        Radiology. 2019; 292: 365-373https://doi.org/10.1148/radiol.2019181960
        • Walsh SLF
        • Sverzellati N
        • Devaraj A
        • et al.
        Chronic hypersensitivity pneumonitis: high resolution computed tomography patterns and pulmonary function indices as prognostic determinants.
        Eur Radiol. 2012; https://doi.org/10.1007/s00330-012-2427-0
        • Chung JH
        • Montner SM
        • Adegunsoye A
        • et al.
        CT findings associated with survival in chronic hypersensitivity pneumonitis.
        Eur Radiol. 2017; 27: 5127-5135https://doi.org/10.1007/s00330-017-4936-3
        • Choe J
        • Chae EJ
        • Kim YJ
        • et al.
        Serial changes of CT findings in patients with chronic hypersensitivity pneumonitis: imaging trajectories and predictors of fibrotic progression and acute exacerbation.
        Eur Radiol. 2020; : 1-11https://doi.org/10.1007/s00330-020-07469-2
        • Jacob J
        • Bartholmai BJ
        • Rajagopalan S
        • et al.
        Automated computer-based CT stratification as a predictor of outcome in hypersensitivity pneumonitis.
        Eur Radiol. 2017; https://doi.org/10.1007/s00330-016-4697-4
        • Morisset J
        • Johannson KA
        • Vittinghoff E
        • et al.
        Use of mycophenolate mofetil or azathioprine for the management of chronic hypersensitivity pneumonitis.
        Chest. 2017; 151: 619-625https://doi.org/10.1016/j.chest.2016.10.029
        • Ley B
        • Elicker BM
        • Hartman TE
        • et al.
        Idiopathic pulmonary fibrosis: CT and risk of death.
        Radiology. 2014; https://doi.org/10.1148/radiol.14130216
        • Raghu G
        • Collard HR
        • Egan JJ
        • et al.
        An official ATS/ERS/JRS/ALAT statement: idiopathic pulmonary fibrosis: evidence-based guidelines for diagnosis and management.
        Am J Respir Crit Care Med. 2011; https://doi.org/10.1164/rccm.2009-040GL