Korean Journal of Radiology

The Korean Society of Radiology (대한영상의학회)
권호 표지
DOI QR코드

DOI QR Code

Prediction of Patient Management in COVID-19 Using Deep Learning-Based Fully Automated Extraction of Cardiothoracic CT Metrics and Laboratory Findings

  • Thomas Weikert (Department of Radiology, University Hospital Basel, University of Basel) ;
  • Saikiran Rapaka (Siemens Healthineers) ;
  • Sasa Grbic (Siemens Healthineers) ;
  • Thomas Re (Siemens Healthineers) ;
  • Shikha Chaganti (Siemens Healthineers) ;
  • David J. Winkel (Department of Radiology, University Hospital Basel, University of Basel) ;
  • Constantin Anastasopoulos (Department of Radiology, University Hospital Basel, University of Basel) ;
  • Tilo Niemann (Department of Radiology, Kantonsspital Baden) ;
  • Benedikt J. Wiggli (Department of Infectious Diseases & Infection Control, Kantonsspital Baden) ;
  • Jens Bremerich (Department of Radiology, University Hospital Basel, University of Basel) ;
  • Raphael Twerenbold (Department of Cardiology, University Hospital Basel, University of Basel) ;
  • Gregor Sommer (Department of Radiology, University Hospital Basel, University of Basel) ;
  • Dorin Comaniciu (Siemens Healthineers) ;
  • Alexander W. Sauter (Department of Radiology, University Hospital Basel, University of Basel)
  • Received : 2020.07.02
  • Accepted : 2020.12.23
  • Published : 2021.06.01
Download PDF
⟨ Previous Next ⟩

Abstract

Objective: To extract pulmonary and cardiovascular metrics from chest CTs of patients with coronavirus disease 2019 (COVID-19) using a fully automated deep learning-based approach and assess their potential to predict patient management. Materials and Methods: All initial chest CTs of patients who tested positive for severe acute respiratory syndrome coronavirus 2 at our emergency department between March 25 and April 25, 2020, were identified (n = 120). Three patient management groups were defined: group 1 (outpatient), group 2 (general ward), and group 3 (intensive care unit [ICU]). Multiple pulmonary and cardiovascular metrics were extracted from the chest CT images using deep learning. Additionally, six laboratory findings indicating inflammation and cellular damage were considered. Differences in CT metrics, laboratory findings, and demographics between the patient management groups were assessed. The potential of these parameters to predict patients' needs for intensive care (yes/no) was analyzed using logistic regression and receiver operating characteristic curves. Internal and external validity were assessed using 109 independent chest CT scans. Results: While demographic parameters alone (sex and age) were not sufficient to predict ICU management status, both CT metrics alone (including both pulmonary and cardiovascular metrics; area under the curve [AUC] = 0.88; 95% confidence interval [CI] = 0.79-0.97) and laboratory findings alone (C-reactive protein, lactate dehydrogenase, white blood cell count, and albumin; AUC = 0.86; 95% CI = 0.77-0.94) were good classifiers. Excellent performance was achieved by a combination of demographic parameters, CT metrics, and laboratory findings (AUC = 0.91; 95% CI = 0.85-0.98). Application of a model that combined both pulmonary CT metrics and demographic parameters on a dataset from another hospital indicated its external validity (AUC = 0.77; 95% CI = 0.66-0.88). Conclusion: Chest CT of patients with COVID-19 contains valuable information that can be accessed using automated image analysis. These metrics are useful for the prediction of patient management.

Keywords

Acknowledgement

We want to thank Ullaskrishnan Poikavilla from Siemens Healthineers, USA, for installing the algorithm prototype at our medical center. Additionally, we appreciate the great support of our research team, namely Rita Achermann, Ivan Nesic, Joshy Cyriac, and Bram Stieltjes.

References

  1. Johns Hopkins University. COVID-19 Map - Johns Hopkins Coronavirus Resource Center. Coronavirus.jhu.edu Web site. https://coronavirus.jhu.edu/map.html. Accessed May 26, 2020
  2. Epidemiology Working Group for NCIP Epidemic Response, Chinese Center for Disease Control and Prevention. The epidemiological characteristics of an outbreak of 2019 novel coronavirus diseases (COVID-19) in China. Zhonghua Liu Xing Bing Xue Za Zhi 2020;41:145-151
  3. Verity R, Okell LC, Dorigatti I, Winskill P, Whittaker C, Imai N, et al. Estimates of the severity of coronavirus disease 2019: a model-based analysis. Lancet Infect Dis 2020;20:669-677
  4. Mehta P, McAuley DF, Brown M, Sanchez E, Tattersall RS, Manson JJ; HLH Across Speciality Collaboration, UK. COVID-19: consider cytokine storm syndromes and immunosuppression. Lancet 2020;395:1033-1034
  5. American College of Radiology. ACR recommendations for the use of chest radiography and computed tomography (CT) for suspected COVID-19 infection. Acr.org Web site. https://www.acr.org/Advocacy-and-Economics/ACR-Position-Statements/Recommendations-for-Chest-Radiography-and-CT-for-Suspected-COVID19-Infection. Published 2020. Accessed August 6, 2020
  6. Hwang EJ, Kim H, Yoon SH, Goo JM, Park CM. Implementation of a deep learning-based computer-aided detection system for the interpretation of chest radiographs in patients suspected for COVID-19. Korean J Radiol 2020;21:1150-1160
  7. Sun D, Li X, Guo D, Wu L, Chen T, Fang Z, et al. CT quantitative analysis and its relationship with clinical features for assessing the severity of patients with COVID-19. Korean J Radiol 2020;21:859-868
  8. Li K, Fang Y, Li W, Pan C, Qin P, Zhong Y, et al. CT image visual quantitative evaluation and clinical classification of coronavirus disease (COVID-19). Eur Radiol 2020;30:4407-4416
  9. Zhang R, Ouyang H, Fu L, Wang S, Han J, Huang K, et al. CT features of SARS-CoV-2 pneumonia according to clinical presentation: a retrospective analysis of 120 consecutive patients from Wuhan city. Eur Radiol 2020;30:4417-4426
  10. Liu Z, Jin C, Wu CC, Liang T, Zhao H, Wang Y, et al. Association between initial chest CT or clinical features and clinical course in patients with coronavirus disease 2019 pneumonia. Korean J Radiol 2020;21:736-745
  11. Wan S, Li M, Ye Z, Yang C, Cai Q, Duan S, et al. CT manifestations and clinical characteristics of 1115 patients with coronavirus disease 2019 (COVID-19): a systematic review and meta-analysis. Acad Radiol 2020;27:910-921
  12. Li M, Lei P, Zeng B, Li Z, Yu P, Fan B, et al. Coronavirus disease (COVID-19): spectrum of CT findings and temporal progression of the disease. Acad Radiol 2020;27:603-608
  13. Zheng Y, Xiao A, Yu X, Zhao Y, Lu Y, Li X, et al. Development and validation of a prognostic nomogram based on clinical and CT features for adverse outcome prediction in patients with COVID-19. Korean J Radiol 2020;21:1007-1017
  14. Yin X, Min X, Nan Y, Feng Z, Li B, Cai W, et al. Assessment of the severity of coronavirus disease: quantitative computed tomography parameters versus semiquantitative visual score. Korean J Radiol 2020;21:998-1006
  15. Park B, Park J, Lim JK, Shin KM, Lee J, Seo H, et al. Prognostic implication of volumetric quantitative ct analysis in patients with COVID-19: a multicenter study in Daegu, Korea. Korean J Radiol 2020;21:1256-1264
  16. Driggin E, Madhavan MV, Bikdeli B, Chuich T, Laracy J, Biondi-Zoccai G, et al. Cardiovascular considerations for patients, health care workers, and health systems during the COVID-19 pandemic. J Am Coll Cardiol 2020;75:2352-2371
  17. Chaganti S, Grenier P, Balachandran A, Chabin G, Cohen S, Flohr T, et al. Automated quantification of CT patterns associated with COVID-19 from chest CT. Radiology: Artificial Intelligence 2020;2:e200048
  18. Ali A, Balachandran A, Vishwanath RS, Barthur A, Wichmann JL, Cimen S, et al. Evaluation of a deep learning based aortic diameter quantification system against multi-reader consensus measurement. Proceedings 2020 European Congress of Radiology (ECR); 2020 Jul 15-19; Vienna, Austria: European Society of Radiology
  19. Martin SS, van Assen M, Rapaka S, Hudson HT Jr, Fischer AM, Varga-Szemes A, et al. Evaluation of a deep learning-based automated CT coronary artery calcium scoring algorithm. JACC Cardiovasc Imaging 2020;13:524-526
  20. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 1988;44:837-845
  21. Tan C, Huang Y, Shi F, Tan K, Ma Q, Chen Y, et al. C-reactive protein correlates with computed tomographic findings and predicts severe COVID-19 early. J Med Virol 2020;92:856-862
  22. Lyu P, Liu X, Zhang R, Shi L, Gao J. The performance of chest CT in evaluating the clinical severity of COVID-19 pneumonia: identifying critical cases based on CT characteristics. Invest Radiol 2020;55:412-421
  23. Francone M, Iafrate F, Masci GM, Coco S, Cilia F, Manganaro L, et al. Chest CT score in COVID-19 patients: correlation with disease severity and short-term prognosis. Eur Radiol 2020;30:6808-6817
  24. Li B, Yang J, Zhao F, Zhi L, Wang X, Liu L, et al. Prevalence and impact of cardiovascular metabolic diseases on COVID-19 in China. Clin Res Cardiol 2020;109:531-538
  25. Ruch Y, Kaeuffer C, Ohana M, Labani A, Fabacher T, Bilbault P, et al. CT lung lesions as predictors of early death or ICU admission in COVID-19 patients. Clin Microbiol Infect 2020;26:1417.e5-1417.e8
  26. Kim YC, Chung Y, Choe YH. Automatic localization of anatomical landmarks in cardiac MR perfusion using random forests. Biomed Signal Proces 2017;38:370-378
  27. Huang L, Han R, Ai T, Yu P, Kang H, Tao Q, et al. Serial quantitative chest CT assessment of COVID-19: a deep learning approach. Radiology: Cardiothoracic Imaging 2020;2:e200075
  28. National Health Commission & National Administration of Traditional Chinese Medicine. Diagnosis and treatment protocol for novel coronavirus pneumonia. Chin Med J (Engl) 2020;133:1087-1095
  29. Zheng YY, Ma YT, Zhang JY, Xie X. COVID-19 and the cardiovascular system. Nat Rev Cardiol 2020;17:259-260