Prediction of lung cancer risk at follow-up screening with low-dose CT: a training and validation study of a deep learning method

Peng Huang, PhD; Cheng T Lin, MD; Yuliang Li, MS; Martin C Tammemagi, ProfPhD; Malcolm V Brock, ProfMD; Sukhinder Atkar-Khattra, BSc; Yanxun Xu, PhD; Ping Hu, ScD; John R Mayo, ProfMD; Heidi Schmidt, ProfMD; Michel Gingras, MD; Sergio Pasian, MD; Lori Stewart, MD; Scott Tsai, MD; Jean M Seely, MD; Daria Manos, MD; Paul Burrowes, MD; Rick Bhatia, MD; Ming-Sound Tsao, ProfMD; Stephen Lam, ProfMD

The Lancet: Digital Health (Nov 2019)

Prediction of lung cancer risk at follow-up screening with low-dose CT: a training and validation study of a deep learning method

Peng Huang, PhD,
Cheng T Lin, MD,
Yuliang Li, MS,
Martin C Tammemagi, ProfPhD,
Malcolm V Brock, ProfMD,
Sukhinder Atkar-Khattra, BSc,
Yanxun Xu, PhD,
Ping Hu, ScD,
John R Mayo, ProfMD,
Heidi Schmidt, ProfMD,
Michel Gingras, MD,
Sergio Pasian, MD,
Lori Stewart, MD,
Scott Tsai, MD,
Jean M Seely, MD,
Daria Manos, MD,
Paul Burrowes, MD,
Rick Bhatia, MD,
Ming-Sound Tsao, ProfMD,
Stephen Lam, ProfMD

Affiliations

Peng Huang, PhD: Department of Oncology, Johns Hopkins University, Baltimore, MD, USA; Department of Biostatistics, Johns Hopkins University, Baltimore, MD, USA; Correspondence to: Dr Peng Huang, Department of Oncology, Johns Hopkins University, Baltimore, MD 21205, USA
Cheng T Lin, MD: Department of Radiology, Johns Hopkins University, Baltimore, MD, USA
Yuliang Li, MS: Department of Applied Mathematics and Statistics, Johns Hopkins University, Baltimore, MD, USA
Martin C Tammemagi, ProfPhD: Department of Community Health Sciences, Brock University, St Catharines, ON, Canada
Malcolm V Brock, ProfMD: Department of Surgery, Johns Hopkins University, Baltimore, MD, USA
Sukhinder Atkar-Khattra, BSc: British Columbia Cancer Agency, Vancouver, BC, Canada
Yanxun Xu, PhD: Department of Applied Mathematics and Statistics, Johns Hopkins University, Baltimore, MD, USA
Ping Hu, ScD: Division of Cancer Prevention, National Cancer Institute, Bethesda, MD, USA
John R Mayo, ProfMD: University of British Columbia and Vancouver General Hospital, Vancouver, BC, Canada
Heidi Schmidt, ProfMD: University Health Network—Princess Margaret Cancer Centre and Toronto General Hospital, Toronto, ON, Canada
Michel Gingras, MD: University Institute of Cardiology and Pneumology, Quebec City, QC, Canada
Sergio Pasian, MD: University Institute of Cardiology and Pneumology, Quebec City, QC, Canada
Lori Stewart, MD: Department of Diagnostic Imaging, Juravinski Hospital, Hamilton, ON, Canada
Scott Tsai, MD: Department of Diagnostic Imaging, Juravinski Hospital, Hamilton, ON, Canada
Jean M Seely, MD: Ottawa Hospital Research Institute and the University of Ottawa, Ottawa, ON, Canada
Daria Manos, MD: Dalhousie University, Halifax, NS, Canada
Paul Burrowes, MD: University of Calgary, Foothills Medical Centre, Calgary, AB, Canada
Rick Bhatia, MD: Memorial University of Newfoundland, St John's, NL, Canada
Ming-Sound Tsao, ProfMD: University Health Network—Princess Margaret Cancer Centre and Toronto General Hospital, Toronto, ON, Canada
Stephen Lam, ProfMD: University of British Columbia—British Columbia Cancer Agency and Vancouver General Hospital, Vancouver, BC, Canada

Journal volume & issue: Vol. 1, no. 7
pp. e353 – e362

Abstract

Read online

Summary: Background: Current lung cancer screening guidelines use either mean diameter, volume, or density of the largest lung nodule on the previous CT scan or appearance of a new nodule to ascertain the timing of the next CT scan. We aimed to develop an accurate screening protocol by estimating the 3-year lung cancer risk after two screening CT scans using deep learning of radiologists' CT readings and other universally available clinical information. Methods: A deep learning algorithm (referred to as DeepLR) was developed using data from participants who had received at least two CT screening scans up to 2 years apart in the National Lung Screening Trial (NLST; training cohort). Double-blinded validation was done using data from participants in the Pan-Canadian Early Detection of Lung Cancer (PanCan) study (validation cohort). The primary analysis was to compare accuracy of DeepLR scores to predict lung cancer incidence at 1 year, 2 years, and 3 years with the Lung CT Screening Reporting & Data System (Lung-RADS) and volume doubling time, using time-dependent area under the receiver operating characteristic curve (AUC) analysis. Findings: The training cohort consisted of 25 097 participants from NLST and the validation cohort comprised 2294 individuals from PanCan. In the validation cohort, DeepLR showed good discrimination, with 1-year, 2-year, and 3-year time-dependent AUC values for cancer diagnosis of 0·968 (SD 0·013), 0·946 (0·013), and 0·899 (0·017), respectively. Among individuals deemed high risk by DeepLR, 94%, 85%, and 71% of incident and interval lung cancers diagnosed within 1 year, 2 years, and 3 years, respectively, after the second screening CT scan were identified. Furthermore, individuals with high DeepLR scores had a significantly higher risk of mortality (hazard ratio 16·07, 95% CI 10·15–25·44; p<0·0001) among people with high scores on Lung-RADS. Interpretation: DeepLR recognises patterns in both temporal and spatial changes and synergy among changes in nodule and non-nodule features. DeepLR scores could be used to accurately guide clinical management after the next scheduled repeat screening CT scan. Funding: Allegheny Health Network, Johns Hopkins University, Terry Fox Research Institute, and British Columbia Cancer Foundation.

Published in The Lancet: Digital Health

ISSN: 2589-7500 (Online)
Publisher: Elsevier
Country of publisher: United Kingdom
LCC subjects: Medicine: Medicine (General): Computer applications to medicine. Medical informatics
Website: https://www.thelancet.com/journals/landig/home

About the journal