Evaluation of the prospects for using artificial intelligence technologies to analyze CT scans of the chest organs in order to identify signs of malignant neoplasms in the lungs

The purpose of the study. To analyze the possibility of using artificial intelligence as a decision support system for radiologists for pulmonary nodules detection on Chest CT before and during the COVID-19 pandemic on the example of the system Botkin.AI.

Materials and methods. Two groups of Chest CT studies were identified: those performed before (group 1) and during the COVID-19 pandemic (group 2). Each group contains anonymized CT data of 150 patients. Chest CT scans for group 2 were selected based on the percentage of coronavirus lung damage from 0 to 25%. The research was analyzed by the artificial intelligence system Botkin. AI for the presence of peripheral pulmonary nodes up to 6 mm, followed by a “blind” check of the analysis results by three radiologists.

Results. In group 1, the sensitivity of the method was 1.0; specificity – 0.88 and AUC – 0.94. In the 2nd group 0.93; 0.81 and 0.86, respectively.

In group 2, a slight decrease in specificity is mainly associated with an increase in false positive results in the pulmonary opcities, as manifestations of coronavirus lung damage, taken by the AI model for pulmonary nodes.

Conclusion. The platform has a high accuracy of detecting pulmonary nodules on computed tomography of the chest both in studies conducted before and during the COVID-19 pandemic. It can be useful to prevent possible omissions of important findings in conditions of increased workload for radiologists.

1. MacMahon H., Naidich D., Goo J., et al. Guidelines for Management of Incidental Pulmonary Nodules Detected on CT Images: From the Fleischner Society 2017. Radiology. 2017;284(1):228-243. DOI: 10.1148/radiol.2017161659

2. Kaprin, A.D., Starinsky, V.V., Petrova G.V. Malignant neoplasms in Russia in 2018 (morbidity and mortality). Moscow: MNIOI im. P.A. Herzen; 2019. 250 p. (In Russ). ISBN 978-5-85502-251-3

3. Kaprin A.D., Starinsky V.V. The state of oncological care for the population of Russia in 2019. Moscow: MNIOI im. P.A. Herzen; 2020. 239 p. (In Russ). ISBN 978-5-85502-255-1

4. Sung H, Ferlay J, Siegel R, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA. A Cancer Journal for Clinicians. 2021;71(3):209-249. DOI: 10.3322/caac.21660

5. Bruno M, Walker E, Abujudeh H. Understanding and Confronting Our Mistakes: The Epidemiology of Error in Radiology and Strategies for Error Reduction. RadioGraphics. 2015;35(6):1668-1676.DOI: 10.1148/rg.2015150023

6. Shi J, Giess C, Martin T, et al. Radiology Workload Changes During the COVID-19 Pandemic: Implications for Staff Redeployment. Academic Radiology. 2021;28(1):1-7.DOI: 10.1016/j.acra.2020.09.008

7. Pinsky P. Lung cancer screening with low-dose CT: a world-wide view. Translational Lung Cancer Research. 2018;7(3):234-242.DOI: 10.21037/tlcr.2018.05.12

8. Liu B, Quan X, Xu C, Lv J, Li C, Dong L, et al. Lung cancer in young adults aged 35 years or younger: A full-scale analysis and review. Journal of Cancer. 2019;10(15):3553-3559.DOI: 10.7150/jca.27490

9. Strand T, Malayeri C, Eskonsipo P, et al. Adolescent smoking and trends in lung cancer incidence among young adults in Norway 1954–1998. Cancer Causes & Control. 2004;15(1):27-33.DOI: 10.1023/b:caco.0000016575.31651.b0

10. Kilsdonk I, de Roos M, Bresser P, Reesink H, Peringa J. Frequency and spectrum of incidental findings when using chest CT as a primary triage tool for COVID-19. European Journal of Radiology Open. 2021; 8:100366. DOI: 10.1016/j.ejro.2021.100366

11. Drokin I, Ericheva E. End-to-End Lung Nodule Detection Framework with Model-Based Feature Projection Block. Machine Learning in Medical Imaging. 2021;12966:91-100. DOI: 10.1007/978-3-030-87589-3_10

12. Drokin I, Ericheva E. Deep Learning on Point Clouds for False Positive Reduction at Nodule Detection in Chest CT Scans. Lecture Notes in Computer Science. 2021;12602:201-215. DOI: 10.1007/978-3-030-72610-2_15

13. Bankier A, MacMahon H, Goo J, et al. Recommendations for Measuring Pulmonary Nodules at CT: A Statement from the Fleischner Society. Radiology. 2017;285(2):584-600. DOI: 10.1148/radiol.2017162894.

14. Kazerooni E, Aberle D, Black W. Lung-RADS v1.1 Assessment Categories Release date: 2019. American College of Radiology.

15. Pedregosa F., Gaël V., Gramfort A. et al. Scikit-learn: Machine Learning in Python. The Journal of Machine Learning Research. 12(85):2825−2830, 2011

16. Morozov S.P., Vladzimirsky A.V., Klyashtorny V.G. Clinical trials of software based on intelligent technologies (specialization "Radiology") Moscow, 2019

17. Li R., Xiao C., Huang Y. et al. Deep Learning Applications in Computed Tomography Images for Pulmonary Nodule Detection and Diagnosis: A Review. Diagnostics (Basel). 2022; 12 (2): 298. https://doi.org/10.3390/diagnostics12020298

18. Setio A.A.A., Traverso A., de Bel T. et al. Validation, comparison, and combination of algorithms for automatic detection of pulmonary nodules in computed tomography images: The LUNA16 challenge. Med. Image Anal. 2017; 42: 1–13. https://doi.org/10.1016/j.media.2017.06.015

19. Kooi T., Litjens G., van Ginneken B. et al. Large scale deep learning for computer aided detection of mammographic lesions. Med. Image Analysis. 2017; 35: 303–312. https://doi.org/10.1016/j.media.2016.07.007

20. Zhang Y., Yang W., Dong Liu et al. COVID-19 and earlystage lung cancer both featuring ground-glass opacities: a propensity score-matched study. Translat. Lung Cancer Res. 2020; 9 (4): 1516–1527. https://doi.org/10.21037/tlcr-20-892

21. Perdigón Martinelli C., Morell C., González C., NovaLozano C. Metastatic pulmonary dissemination as differential diagnosis of COVID-19 disease. BMJ Case Rep. 2020; 13 (10): e237453. https://doi.org/10.1136/bcr-2020237453