Matthew Moll1, Dandi Qiao2, Elizabeth A Regan3, Gary M Hunninghake4, Barry J Make5, Ruth Tal-Singer6, Michael J McGeachie2, Peter J Castaldi2, Raul San Jose Estepar7, George R Washko7, James M Wells8, David LaFon8, Matthew Strand5, Russell P Bowler9, MeiLan K Han10, Jorgen Vestbo11, Bartolome Celli4, Peter Calverley12, James Crapo5, Edwin K Silverman1, Brian D Hobbs1, Michael H Cho13. 1. Channing Division of Network Medicine, Brigham and Women's Hospital, Boston, MA; Division of Pulmonary and Critical Care Medicine, Brigham and Women's Hospital, Boston, MA. 2. Channing Division of Network Medicine, Brigham and Women's Hospital, Boston, MA. 3. Division of Pulmonary and Critical Care Medicine, University of Colorado, Denver, CO. 4. Division of Pulmonary and Critical Care Medicine, Brigham and Women's Hospital, Boston, MA. 5. Division of Pulmonary and Critical Care Medicine, National Jewish Health, Denver, CO. 6. GlaxoSmithKline Research and Development, Collegeville, PA. 7. Division of Pulmonary and Critical Care Medicine, Brigham and Women's Hospital, Boston, MA; Applied Chest Imaging Laboratory, Brigham and Women's Hospital, Boston, MA. 8. Division of Pulmonary, Allergy and Critical Care Medicine, University of Alabama at Birmingham, Birmingham, AL. 9. Division of Pulmonary and Critical Care Medicine, University of Colorado, Denver, CO; Division of Pulmonary and Critical Care Medicine, National Jewish Health, Denver, CO. 10. Division of Pulmonary and Critical Care Medicine, University of Michigan Health System, Ann Arbor, MI. 11. Division of Infection, Immunity and Respiratory Medicine, Manchester Academic Health Sciences Centre, The University of Manchester and the Manchester University NHS Foundation Trust, Manchester, England. 12. Department of Medicine, University of Liverpool, Liverpool, England. 13. Channing Division of Network Medicine, Brigham and Women's Hospital, Boston, MA; Division of Pulmonary and Critical Care Medicine, Brigham and Women's Hospital, Boston, MA. Electronic address: remhc@channing.harvard.edu.
Abstract
BACKGROUND: COPD is a leading cause of mortality. RESEARCH QUESTION: We hypothesized that applying machine learning to clinical and quantitative CT imaging features would improve mortality prediction in COPD. STUDY DESIGN AND METHODS: We selected 30 clinical, spirometric, and imaging features as inputs for a random survival forest. We used top features in a Cox regression to create a machine learning mortality prediction (MLMP) in COPD model and also assessed the performance of other statistical and machine learning models. We trained the models in subjects with moderate to severe COPD from a subset of subjects in Genetic Epidemiology of COPD (COPDGene) and tested prediction performance in the remainder of individuals with moderate to severe COPD in COPDGene and Evaluation of COPD Longitudinally to Identify Predictive Surrogate Endpoints (ECLIPSE). We compared our model with the BMI, airflow obstruction, dyspnea, exercise capacity (BODE) index; BODE modifications; and the age, dyspnea, and airflow obstruction index. RESULTS: We included 2,632 participants from COPDGene and 1,268 participants from ECLIPSE. The top predictors of mortality were 6-min walk distance, FEV1 % predicted, and age. The top imaging predictor was pulmonary artery-to-aorta ratio. The MLMP-COPD model resulted in a C index ≥ 0.7 in both COPDGene and ECLIPSE (6.4- and 7.2-year median follow-ups, respectively), significantly better than all tested mortality indexes (P < .05). The MLMP-COPD model had fewer predictors but similar performance to that of other models. The group with the highest BODE scores (7-10) had 64% mortality, whereas the highest mortality group defined by the MLMP-COPD model had 77% mortality (P = .012). INTERPRETATION: An MLMP-COPD model outperformed four existing models for predicting all-cause mortality across two COPD cohorts. Performance of machine learning was similar to that of traditional statistical methods. The model is available online at: https://cdnm.shinyapps.io/cgmortalityapp/.
BACKGROUND:COPD is a leading cause of mortality. RESEARCH QUESTION: We hypothesized that applying machine learning to clinical and quantitative CT imaging features would improve mortality prediction in COPD. STUDY DESIGN AND METHODS: We selected 30 clinical, spirometric, and imaging features as inputs for a random survival forest. We used top features in a Cox regression to create a machine learning mortality prediction (MLMP) in COPD model and also assessed the performance of other statistical and machine learning models. We trained the models in subjects with moderate to severe COPD from a subset of subjects in Genetic Epidemiology of COPD (COPDGene) and tested prediction performance in the remainder of individuals with moderate to severe COPD in COPDGene and Evaluation of COPD Longitudinally to Identify Predictive Surrogate Endpoints (ECLIPSE). We compared our model with the BMI, airflow obstruction, dyspnea, exercise capacity (BODE) index; BODE modifications; and the age, dyspnea, and airflow obstruction index. RESULTS: We included 2,632 participants from COPDGene and 1,268 participants from ECLIPSE. The top predictors of mortality were 6-min walk distance, FEV1 % predicted, and age. The top imaging predictor was pulmonary artery-to-aorta ratio. The MLMP-COPD model resulted in a C index ≥ 0.7 in both COPDGene and ECLIPSE (6.4- and 7.2-year median follow-ups, respectively), significantly better than all tested mortality indexes (P < .05). The MLMP-COPD model had fewer predictors but similar performance to that of other models. The group with the highest BODE scores (7-10) had 64% mortality, whereas the highest mortality group defined by the MLMP-COPD model had 77% mortality (P = .012). INTERPRETATION: An MLMP-COPD model outperformed four existing models for predicting all-cause mortality across two COPD cohorts. Performance of machine learning was similar to that of traditional statistical methods. The model is available online at: https://cdnm.shinyapps.io/cgmortalityapp/.
Authors: Natalie Terzikhan; Daniel Bos; Lies Lahousse; Lennard Wolff; Katia M C Verhamme; Maarten J G Leening; Janine F Felix; Henning Gall; Hossein A Ghofrani; Oscar H Franco; M Arfan Ikram; Bruno H Stricker; Aad van der Lugt; Guy Brusselle Journal: Eur Respir J Date: 2017-06-15 Impact factor: 16.671
Authors: Joy Miller; Lisa D Edwards; Alvar Agustí; Per Bakke; Peter M A Calverley; Bartolome Celli; Harvey O Coxson; Courtney Crim; David A Lomas; Bruce E Miller; Steve Rennard; Edwin K Silverman; Ruth Tal-Singer; Jørgen Vestbo; Emiel Wouters; Julie C Yates; William Macnee Journal: Respir Med Date: 2013-06-19 Impact factor: 3.415
Authors: Jose M Marin; Inmaculada Alfageme; Pere Almagro; Ciro Casanova; Cristobal Esteban; Juan J Soler-Cataluña; Juan P de Torres; Pablo Martínez-Camblor; Marc Miravitlles; Bartolome R Celli; Joan B Soriano Journal: Eur Respir J Date: 2012-12-06 Impact factor: 16.671
Authors: Andrew H Briggs; Timothy Baker; Nancy A Risebrough; Mike Chambers; Sebastian Gonzalez-McQuire; Afisi S Ismaila; Alex Exuzides; Chris Colby; Maggie Tabberer; Hana Muellerova; Nicholas Locantore; Maureen P M H Rutten van Mölken; David A Lomas Journal: Med Decis Making Date: 2016-06-17 Impact factor: 2.583
Authors: Bartolome R Celli; Nicholas Locantore; Julie Yates; Ruth Tal-Singer; Bruce E Miller; Per Bakke; Peter Calverley; Harvey Coxson; Courtney Crim; Lisa D Edwards; David A Lomas; Annelyse Duvoix; William MacNee; Stephen Rennard; Edwin Silverman; Jørgen Vestbo; Emiel Wouters; Alvar Agustí Journal: Am J Respir Crit Care Med Date: 2012-03-15 Impact factor: 21.405
Authors: Beniamino Guerra; Sarah R Haile; Bernd Lamprecht; Ana S Ramírez; Pablo Martinez-Camblor; Bernhard Kaiser; Inmaculada Alfageme; Pere Almagro; Ciro Casanova; Cristóbal Esteban-González; Juan J Soler-Cataluña; Juan P de-Torres; Marc Miravitlles; Bartolome R Celli; Jose M Marin; Gerben Ter Riet; Patricia Sobradillo; Peter Lange; Judith Garcia-Aymerich; Josep M Antó; Alice M Turner; Meilan K Han; Arnulf Langhammer; Linda Leivseth; Per Bakke; Ane Johannessen; Toru Oga; Borja Cosio; Julio Ancochea-Bermúdez; Andres Echazarreta; Nicolas Roche; Pierre-Régis Burgel; Don D Sin; Joan B Soriano; Milo A Puhan Journal: BMC Med Date: 2018-03-02 Impact factor: 8.775
Authors: Michael I Polkey; Martijn A Spruit; Lisa D Edwards; Michael L Watkins; Victor Pinto-Plata; Jørgen Vestbo; Peter M A Calverley; Ruth Tal-Singer; Alvar Agustí; Per S Bakke; Harvey O Coxson; David A Lomas; William MacNee; Stephen Rennard; Edwin K Silverman; Bruce E Miller; Courtney Crim; Julie Yates; Emiel F M Wouters; Bartolome Celli Journal: Am J Respir Crit Care Med Date: 2012-12-21 Impact factor: 21.405
Authors: Matthew Moll; Adel Boueiz; Auyon J Ghosh; Aabida Saferali; Sool Lee; Zhonghui Xu; Jeong H Yun; Brian D Hobbs; Craig P Hersh; Don D Sin; Ruth Tal-Singer; Edwin K Silverman; Michael H Cho; Peter J Castaldi Journal: Am J Respir Crit Care Med Date: 2022-01-15 Impact factor: 21.405
Authors: Salvatore Tedesco; Martina Andrulli; Markus Åkerlund Larsson; Daniel Kelly; Antti Alamäki; Suzanne Timmons; John Barton; Joan Condell; Brendan O'Flynn; Anna Nordström Journal: Int J Environ Res Public Health Date: 2021-12-04 Impact factor: 3.390