CoChat
Sign Up
Cited by 140Radiomics and Machine Learning in Medical ImagingAI in cancer detectionMachine Learning in Healthcare

Deep Learning to Assess Long-term Mortality From Chest Radiographs

Open access
Lu, Michael T., Ivanov, Alexander, Mayrhofer, Thomas, Hosny, Ahmed, Aerts, Hugo J. W. L., Hoffmann, Udo·2019·JAMA Network Open
DOI 10.1001/jamanetworkopen.2019.7416PMC PMC6646994
Open this paper in CoChat — free accountOpen PDF

Abstract

Importance: Chest radiography is the most common diagnostic imaging test in medicine and may also provide information about longevity and prognosis. Objective: To develop and test a convolutional neural network (CNN) (named CXR-risk) to predict long-term mortality, including noncancer death, from chest radiographs. Design, Setting, and Participants: In this prognostic study, CXR-risk CNN development (n = 41 856) and testing (n = 10 464) used data from the screening radiography arm of the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO) (n = 52 320), a community cohort of asymptomatic nonsmokers and smokers (aged 55-74 years) enrolled at 10 US sites from November 8, 1993, through July 2, 2001. External testing used data from the screening radiography arm of the National Lung Screening Trial (NLST) (n = 5493), a community cohort of heavy smokers (aged 55-74 years) enrolled at 21 US sites from August 2002, through April 2004. Data analysis was performed from January 1, 2018, to May 23, 2019. Exposure: Deep learning CXR-risk score (very low, low, moderate, high, and very high) based on CNN analysis of the enrollment radiograph. Main Outcomes and Measures: All-cause mortality. Prognostic value was assessed in the context of radiologists' diagnostic findings (eg, lung nodule) and standard risk factors (eg, age, sex, and diabetes) and for cause-specific mortality. Results: Among 10 464 PLCO participants (mean [SD] age, 62.4 [5.4] years; 5405 men [51.6%]; median follow-up, 12.2 years [interquartile range, 10.5-12.9 years]) and 5493 NLST test participants (mean [SD] age, 61.7 [5.0] years; 3037 men [55.3%]; median follow-up, 6.3 years [interquartile range, 6.0-6.7 years]), there was a graded association between CXR-risk score and mortality. The very high-risk group had mortality of 53.0% (PLCO) and 33.9% (NLST), which was higher compared with the very low-risk group (PLCO: unadjusted hazard ratio [HR], 18.3 [95% CI, 14.5-23.2]; NLST: unadjusted HR, 15.2 [95% CI, 9.2-25.3]; both P < .001). This association was robust to adjustment for radiologists' findings and risk factors (PLCO: adjusted HR [aHR], 4.8 [95% CI, 3.6-6.4]; NLST: aHR, 7.0 [95% CI, 4.0-12.1]; both P < .001). Comparable results were seen for lung cancer death (PLCO: aHR, 11.1 [95% CI, 4.4-27.8]; NLST: aHR, 8.4 [95% CI, 2.5-28.0]; both P ≤ .001) and for noncancer cardiovascular death (PLCO: aHR, 3.6 [95% CI, 2.1-6.2]; NLST: aHR, 47.8 [95% CI, 6.1-374.9]; both P < .001) and respiratory death (PLCO: aHR, 27.5 [95% CI, 7.7-97.8]; NLST: aHR, 31.9 [95% CI, 3.9-263.5]; both P ≤ .001). Conclusions and Relevance: In this study, the deep learning CXR-risk score stratified the risk of long-term mortality based on a single chest radiograph. Individuals at high risk of mortality may benefit from prevention, screening, and lifestyle interventions.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1001/jamanetworkopen.2019.7416

Cites

10 of 10 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Deep learning can capture imaging features beyond traditional findings (e.g. enlarged heart silhouette, lung nodules) with valuable prognostic and potentially actionable information.

    “the results of this and previously published studies(10–12) suggest that deep learning can capture imaging features beyond traditional findings (e.g. enlarged heart silhouette, lung nodules) with valuable prognostic and potentially actionable information.”

    From Deep Learning to Estimate Cardiovascular Risk From Chest Radiographs (2024) · cites this paper as [10]

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

Research works printed but not linked: 41

  1. [1]Ron E Cancer risks from medical radiation. Health Phys. 2003; 85(1):-. doi: 10.1097/00004032-200307000-00011 12852471 · DOI · Paper page
  2. [2]Rosman DA, Duszak R Jr, Wang W, Hughes DR, Rosenkrantz AB Changing utilization of noninvasive diagnostic imaging over 2 decades: an examination family-focused analysis of Medicare claims using the Neiman Imaging Types of Service categorization system. AJR Am J Roentgenol. 2018; 210(2): 364- 368. doi: 10.2214/AJR.17.18214 29220208 · DOI
  3. [3]Bell MF, Jernigan TP, Schaaf RS Prognostic significance of calcification of the aortic knob visualized radiographically. Am J Cardiol. 1964; 13: 640- 644. doi: 10.1016/0002-9149(64)90198-5 14156026 · DOI
  4. [4]Cohn JN, Johnson GR, Shabetai R,; V-HeFT VA Cooperative Studies Group. Ejection fraction, peak exercise oxygen consumption, cardiothoracic ratio, ventricular arrhythmias, and plasma norepinephrine as determinants of prognosis in heart failure. Circulation. 1993; 87(6)(suppl): VI5- VI16. 8500240
  5. [5]Giamouzis G, Sui X, Love TE, Butler J, Young JB, Ahmed A A propensity-matched study of the association of cardiothoracic ratio with morbidity and mortality in chronic heart failure. Am J Cardiol. 2008; 101(3): 343- 347. doi: 10.1016/j.amjcard.2007.08.039 18237597 PMC2659172 · DOI
  6. [6]Olshansky SJ From lifespan to healthspan. JAMA. 2018; 320(13): 1323- 1324. doi: 10.1001/jama.2018.12621 30242384 · DOI
  7. [8]LeCun Y, Bengio Y, Hinton G Deep learning. Nature. 2015; 521(7553): 436- 444. doi: 10.1038/nature14539 26017442 · DOI · Paper page
  8. [9]Hinton G Deep learning—a technology with the potential to transform health care. JAMA. 2018; 320(11): 1101- 1102. doi: 10.1001/jama.2018.11100 30178065 · DOI · Paper page
  9. [10]Wang X, Peng Y, Lu L, Lu Z, Bagheri M, Summers RM ChestX-ray8: hospital-scale chest x-ray database and benchmarks on weakly-supervised classification and localization of common thorax diseases. 2017 IEEE Conference on Computer Vision and Pattern Recognition 2017;2097-2106. http://openaccess.thecvf.com/content_cvpr_2017/html/Wang_ChestX-ray8_Hospital-Scale_Chest_CVPR_2017_paper.html. Accessed May 01, 2017.
  10. [13]Putha P, Tadepalli M, Reddy B, Can artificial intelligence reliably report chest x-rays? radiologist validation of an algorithm trained on 1.2 million x-rays. Preprint. Posted online July 19, 2018. arXiv 1807.07455. · arXiv
  11. [14]Singh R, Kalra MK, Nitiwarangkul C, Deep learning in chest radiography: detection of findings and presence of change. PLoS One. 2018; 13(10): e0204155. doi: 10.1371/journal.pone.0204155 30286097 PMC6171827 · DOI · Paper page
  12. [15]Taylor AG, Mielke C, Mongan J Automated detection of moderate and large pneumothorax on frontal chest X-rays using deep convolutional neural networks: a retrospective study. PLoS Med. 2018; 15(11): e1002697. doi: 10.1371/journal.pmed.1002697 30457991 PMC6245672 · DOI · Paper page
  13. [17]Oken MM, Hocking WG, Kvale PA,; PLCO Project Team Screening by chest radiograph and lung cancer mortality: the Prostate, Lung, Colorectal, and Ovarian (PLCO) randomized trial. JAMA. 2011; 306(17): 1865- 1873. doi: 10.1001/jama.2011.1591 22031728 · DOI
  14. [19]Prorok PC, Andriole GL, Bresalier RS,; Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial Project Team Design of the Prostate, Lung, Colorectal and Ovarian (PLCO) cancer screening trial. Control Clin Trials. 2000; 21(6)(suppl): 273S- 309S. doi: 10.1016/S0197-2456(00)00098-2 11189684 · DOI
  15. [20]Zhu CS, Pinsky PF, Moler JE, Data sharing in clinical trials: an experience with two large cancer screening trials. PLoS Med. 2017; 14(5): e1002304. doi: 10.1371/journal.pmed.1002304 28542201 PMC5441574 · DOI
  16. [21]Aberle DR, Berg CD, Black WC,; National Lung Screening Trial Research Team The National Lung Screening Trial: overview and study design. Radiology. 2011; 258(1): 243- 253. doi: 10.1148/radiol.10091808 21045183 PMC3009383 · DOI
  17. [22]Pinsky PF, Miller A, Kramer BS, Evidence of a healthy volunteer effect in the prostate, lung, colorectal, and ovarian cancer screening trial. Am J Epidemiol. 2007; 165(8): 874- 881. doi: 10.1093/aje/kwk075 17244633 · DOI
  18. [23]Parmar C, Barry JD, Hosny A, Quackenbush J, Aerts HJWL Data analysis strategies in medical imaging. Clin Cancer Res. 2018; 24(15): 3492- 3499. doi: 10.1158/1078-0432.CCR-18-0385 29581134 PMC6082690 · DOI
  19. [24]Szegedy C, Ioffe S, Vanhoucke V, Alemi A Inception-v4, inception-resnet and the impact of residual connections on learning. Preprint. Posted online February 23, 2016. arXiv 1602.07261. · arXiv
  20. [25]Selvaraju RR, Cogswell M, Das A, Vedantam R, Parikh D, Batra D Grad-CAM: visual explanations from deep networks via gradient-based localization. Preprint. Posted online October 7, 2016. arXiv 1610.02391. · arXiv
  21. [27]Grønnesby JK, Borgan O A method for checking regression models in survival analysis based on the risk score. Lifetime Data Anal. 1996; 2(4): 315- 328. doi: 10.1007/BF00127305 9384628 · DOI
  22. [28]DeLong ER, DeLong DM, Clarke-Pearson DL Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics. 1988; 44(3): 837- 845. doi: 10.2307/2531595 3203132 · DOI · Paper page
  23. [30]Pencina MJ, D’Agostino RB Sr, Steyerberg EW Extensions of net reclassification improvement calculations to measure usefulness of new biomarkers. Stat Med. 2011; 30(1): 11- 21. doi: 10.1002/sim.4085 21204120 PMC3341973 · DOI
  24. [31]Steyerberg EW, Vickers AJ, Cook NR, Assessing the performance of prediction models: a framework for traditional and novel measures. Epidemiology. 2010; 21(1): 128- 138. doi: 10.1097/EDE.0b013e3181c30fb2 20010215 PMC3575184 · DOI · Paper page
  25. [32]Poplin R, Varadarajan AV, Blumer K, Prediction of cardiovascular risk factors from retinal fundus photographs via deep learning. Nat Biomed Eng. 2018; 2(3): 158- 164. doi: 10.1038/s41551-018-0195-0 31015713 · DOI · Paper page
  26. [33]González G, Ash SY, Vegas-Sánchez-Ferrero G,; COPDGene and ECLIPSE Investigators Disease staging and prognosis in smokers using deep learning in chest computed tomography. Am J Respir Crit Care Med. 2018; 197(2): 193- 203. doi: 10.1164/rccm.201705-0860OC 28892454 PMC5768902 · DOI
  27. [34]Handy CE, Quispe R, Pinto X, Synergistic opportunities in the interplay between cancer screening and cardiovascular disease risk assessment. Circulation. 2018; 138(7): 727- 734. doi: 10.1161/CIRCULATIONAHA.118.035516 30359131 · DOI
  28. [35]Pursnani A, Massaro JM, D’Agostino RB Sr, O’Donnell CJ, Hoffmann U Guideline-based statin eligibility, cancer events, and noncardiovascular mortality in the Framingham Heart Study. J Clin Oncol. 2017; 35(25): 2927- 2933. doi: 10.1200/JCO.2016.71.3594 28700275 PMC5578389 · DOI
  29. [36]Handy CE, Desai CS, Dardari ZA, The Association of coronary artery calcium with noncardiovascular disease: the multi-ethnic study of atherosclerosis. JACC Cardiovasc Imaging. 2016; 9(5): 568- 576. doi: 10.1016/j.jcmg.2015.09.020 26970999 PMC4860157 · DOI
  30. [37]Ridker PM, MacFadyen JG, Thuren T, Everett BM, Libby P, Glynn RJ; CANTOS Trial Group Effect of interleukin-1β inhibition with canakinumab on incident lung cancer in patients with atherosclerosis: exploratory results from a randomised, double-blind, placebo-controlled trial. Lancet. 2017; 390(10105): 1833- 1842. doi: 10.1016/S0140-6736(17)32247-X 28855077 · DOI · Paper page
  31. [39]Detrano R, Guerci AD, Carr JJ, Coronary calcium as a predictor of coronary events in four racial or ethnic groups. N Engl J Med. 2008; 358(13): 1336- 1345. doi: 10.1056/NEJMoa072100 18367736 · DOI · Paper page
  32. [41]Stone NJ, Robinson JG, Lichtenstein AH,; American College of Cardiology/American Heart Association Task Force on Practice Guidelines 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation. 2014; 129(25)(suppl 2): S1- S45. doi: 10.1161/01.cir.0000437738.63853.7a 24222016 · DOI · Paper page
  33. [43]Moyer VA; U.S. Preventive Services Task Force Screening for lung cancer: US Preventive Services Task Force recommendation statement. Ann Intern Med. 2014; 160(5): 330- 338. doi: 10.7326/M13-2771 24378917 · DOI
  34. [44]Jemal A, Fedewa SA Lung cancer screening with low-dose computed tomography in the United States—2010 to 2015. JAMA Oncol. 2017; 3(9): 1278- 1281. doi: 10.1001/jamaoncol.2016.6416 28152136 PMC5824282 · DOI
  35. [46]Hunter DJ, Drazen JM Has the genome granted our wish yet? N Engl J Med. 2019; 380: 2391- 2393. doi: 10.1056/NEJMp1904511 31091368 · DOI
  36. [47]Emanuel EJ, Wachter RM Artificial intelligence in health care: will the value match the hype? JAMA. 2019; 321(23): 2281- 2282. doi: 10.1001/jama.2019.4914 31107500 · DOI · Paper page
  37. [48]Holzinger A, Biemann C, Pattichis CS, Kell DB What do we need to build explainable AI systems for the medical domain? Preprint. Posted online December 28, 2017. arXiv 1712.9923.
  38. [49]Avati A, Duan T, Jung K, Shah NH, Ng A Countdown regression: sharp and calibrated survival predictions. Preprint. Posted online June 21, 2018. arXiv 1806.08324. · arXiv
  39. [51]Li H, Boimel P, Janopaul-Naylor J, Deep convolutional neural networks for imaging data based survival analysis of rectal cancer. Preprint. Posted online January 5, 2019. arXiv 1901.01449. PMC6955095 31929858 · arXiv
  40. [52]Baltruschat IM, Nickisch H, Grass M, Knopp T, Saalbach A Comparison of deep learning approaches for multi-label chest x-ray classification. Sci Rep. 2019; 9(1): 6381. doi: 10.1038/s41598-019-42294-8 31011155 PMC6476887 · DOI · Paper page
  41. [53]Buolamwini J, Gebru T Gender shades: intersectional accuracy disparities in commercial gender classification. Proc Machine Learning Res. 2018; 81: 77- 91.

Cited by

Sorted by citations

Authors
Lu, M. T., Ivanov, A., Mayrhofer, T., Hosny, A., Aerts, H. J. W. L., & Hoffmann, U.
Affiliations
  • Harvard University (US)
  • Brigham and Women's Hospital (US)
Corresponding author
Lu, Michael T.
ORCID
Venue
JAMA Network Open
Publisher
American Medical Association (AMA)
Published
2019-07-19
Volume
2
Issue
7
Pages
e197416
ISSN
2574-3805 (ISSN-L)
DOI
10.1001/jamanetworkopen.2019.7416
PMC
PMC6646994
OpenAlex
W2963300950
Cited by
140
References
10
Citations by year
2026: 7 · 2025: 11 · 2024: 19 · 2023: 25 · 2022: 26 · 2021: 36 · 2020: 15 · 2019: 1
FWCI
12.65
Type
journal article
Language
en
Open access
Licence
cc-by
Locations
  • JAMA Network Open (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
Funders
  • Brigham and Women's Hospital
  • Massachusetts General Hospital
Topics
  • Health Sciences › Medicine › Radiology, Nuclear Medicine and Imaging › Radiomics and Machine Learning in Medical Imaging
  • Physical Sciences › Computer Science › Artificial Intelligence › AI in cancer detection
  • Physical Sciences › Computer Science › Artificial Intelligence › Machine Learning in Healthcare
Keywords
long-term mortality prediction, chest radiograph, convolutional neural networks, all-cause mortality, lung cancer screening, PLCO trial, prognostic imaging, pulmonary nodules, smoking, asymptomatic adults, risk stratification
MeSH terms
Deep Learning, Aged, Cardiovascular Diseases, Clinical Trials as Topic, Female, Humans, Lung Diseases, Male, Mass Screening, Middle Aged, Prognosis, Risk Factors, Radiography, Thoracic, Thorax, Neural Networks, Computer, Risk Assessment

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review