External validation of a convolutional neural network artificial intelligence tool to predict malignancy in pulmonary nodules
David R Baldwin, Jennifer Gustafson, Lyndsey Pickup, Carlos Arteta, Petr Novotny, Jerome Declerck, Timor Kadir, Catarina Figueiras, Albert Sterba, Alan Exell, Vaclav Potesil, Paul Holland, Hazel Spence, Alison Clubley, Emma O'Dowd, Matthew Clark, Victoria Ashford-Turner, Matthew Ej Callister, Fergus V Gleeson, David R Baldwin, Jennifer Gustafson, Lyndsey Pickup, Carlos Arteta, Petr Novotny, Jerome Declerck, Timor Kadir, Catarina Figueiras, Albert Sterba, Alan Exell, Vaclav Potesil, Paul Holland, Hazel Spence, Alison Clubley, Emma O'Dowd, Matthew Clark, Victoria Ashford-Turner, Matthew Ej Callister, Fergus V Gleeson
Abstract
Background: Estimation of the risk of malignancy in pulmonary nodules detected by CT is central in clinical management. The use of artificial intelligence (AI) offers an opportunity to improve risk prediction. Here we compare the performance of an AI algorithm, the lung cancer prediction convolutional neural network (LCP-CNN), with that of the Brock University model, recommended in UK guidelines.
Methods: A dataset of incidentally detected pulmonary nodules measuring 5-15 mm was collected retrospectively from three UK hospitals for use in a validation study. Ground truth diagnosis for each nodule was based on histology (required for any cancer), resolution, stability or (for pulmonary lymph nodes only) expert opinion. There were 1397 nodules in 1187 patients, of which 234 nodules in 229 (19.3%) patients were cancer. Model discrimination and performance statistics at predefined score thresholds were compared between the Brock model and the LCP-CNN.
Results: The area under the curve for LCP-CNN was 89.6% (95% CI 87.6 to 91.5), compared with 86.8% (95% CI 84.3 to 89.1) for the Brock model (p≤0.005). Using the LCP-CNN, we found that 24.5% of nodules scored below the lowest cancer nodule score, compared with 10.9% using the Brock score. Using the predefined thresholds, we found that the LCP-CNN gave one false negative (0.4% of cancers), whereas the Brock model gave six (2.5%), while specificity statistics were similar between the two models.
Conclusion: The LCP-CNN score has better discrimination and allows a larger proportion of benign nodules to be identified without missing cancers than the Brock model. This has the potential to substantially reduce the proportion of surveillance CT scans required and thus save significant resources.
Keywords: CT imaging; lung cancer; non-small cell lung cancer.
Conflict of interest statement
Competing interests: Several members of the authorship are employed by Optellum, the company that has developed the risk prediction artificial intelligence tool.
© Author(s) (or their employer(s)) 2020. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.
Figures
References
- Callister MEJ, Baldwin DR, Akram AR, et al. . British thoracic Society guidelines for the investigation and management of pulmonary nodules. Thorax 2015;70(Suppl 2):ii1–54. 10.1136/thoraxjnl-2015-207168
- MacMahon H, Naidich DP, Goo JM, et al. . Guidelines for management of incidental pulmonary nodules detected on CT images: from the Fleischner Society 2017. Radiology 2017;284:228–43. 10.1148/radiol.2017161659
- Rami-Porta R, Bolejack V, Crowley J, et al. . The IASLC lung cancer staging project: proposals for the revisions of the T descriptors in the forthcoming eighth edition of the TNM classification for lung cancer. J Thorac Oncol 2015;10:990–1003. 10.1097/JTO.0000000000000559
- McWilliams A, Tammemagi MC, Mayo JR, et al. . Probability of cancer in pulmonary nodules detected on first screening CT. N Engl J Med 2013;369:910–9. 10.1056/NEJMoa1214726
- van Riel SJ, Ciompi F, Jacobs C, et al. . Malignancy risk estimation of screen-detected nodules at baseline CT: comparison of the PanCan model, Lung-RADS and NCCN guidelines. Eur Radiol 2017;27:4019–29. 10.1007/s00330-017-4767-2
- Winkler Wille MM, van Riel SJ, Saghir Z, et al. . Predictive accuracy of the PanCan lung cancer risk prediction model -external validation based on CT from the Danish lung cancer screening trial. Eur Radiol 2015;25:3093–9. 10.1007/s00330-015-3689-0
- Al-Ameri A, Malhotra P, Thygesen H, et al. . Risk of malignancy in pulmonary nodules: a validation study of four prediction models. Lung Cancer 2015;89:27–30. 10.1016/j.lungcan.2015.03.018
- Nair A, Baldwin DR, Field JK, et al. . Measurement methods and algorithms for the management of solid nodules. J Thorac Imaging 2012;27:230–9. 10.1097/RTI.0b013e31824f83e1
- Revel M-P, Bissery A, Bienvenu M, et al. . Are two-dimensional CT measurements of small noncalcified pulmonary nodules reliable? Radiology 2004;231:453–8. 10.1148/radiol.2312030167
- LeCun Y, Bengio Y, Hinton G. Deep learning. Nature 2015;521:436–44. 10.1038/nature14539
- Huang G, Liu Z, Lvd M. , eds Densely connected Convolutional networks. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017. , 21-26 July 2017.
- Aberle DR, Adams AM, Berg CD, et al. . Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med 2011;365:395–409. 10.1056/NEJMoa1102873
- Oke JL, Pickup LC, Declerck J, et al. . Development and validation of clinical prediction models to risk stratify patients presenting with small pulmonary nodules: a research protocol. Diagn Progn Res 2018;2:22 10.1186/s41512-018-0044-3
- Efron BTR. An introduction to the bootstrap. New York: Chapman and Hall, 1993.
- Horeweg N, van Rosmalen J, Heuvelmans MA, et al. . Lung cancer probability in patients with CT-detected pulmonary nodules: a prespecified analysis of data from the Nelson trial of low-dose CT screening. Lancet Oncol 2014;15:1332–41. 10.1016/S1470-2045(14)70389-4
Source: PubMed