Skip to main content
Erschienen in: European Radiology 12/2020

02.07.2020 | COVID-19 | Computed Tomography Zur Zeit gratis

A deep learning approach to characterize 2019 coronavirus disease (COVID-19) pneumonia in chest CT images

verfasst von: Qianqian Ni, Zhi Yuan Sun, Li Qi, Wen Chen, Yi Yang, Li Wang, Xinyuan Zhang, Liu Yang, Yi Fang, Zijian Xing, Zhen Zhou, Yizhou Yu, Guang Ming Lu, Long Jiang Zhang

Erschienen in: European Radiology | Ausgabe 12/2020

Einloggen, um Zugang zu erhalten

Abstract

Objectives

To utilize a deep learning model for automatic detection of abnormalities in chest CT images from COVID-19 patients and compare its quantitative determination performance with radiological residents.

Methods

A deep learning algorithm consisted of lesion detection, segmentation, and location was trained and validated in 14,435 participants with chest CT images and definite pathogen diagnosis. The algorithm was tested in a non-overlapping dataset of 96 confirmed COVID-19 patients in three hospitals across China during the outbreak. Quantitative detection performance of the model was compared with three radiological residents with two experienced radiologists’ reading reports as reference standard by assessing the accuracy, sensitivity, specificity, and F1 score.

Results

Of 96 patients, 88 had pneumonia lesions on CT images and 8 had no abnormities on CT images. For per-patient basis, the algorithm showed superior sensitivity of 1.00 (95% confidence interval (CI) 0.95, 1.00) and F1 score of 0.97 in detecting lesions from CT images of COVID-19 pneumonia patients. While for per-lung lobe basis, the algorithm achieved a sensitivity of 0.96 (95% CI 0.94, 0.98) and a slightly inferior F1 score of 0.86. The median volume of lesions calculated by algorithm was 40.10 cm3. An average running speed of 20.3 s ± 5.8 per case demonstrated the algorithm was much faster than the residents in assessing CT images (all p < 0.017). The deep learning algorithm can also assist radiologists make quicker diagnosis (all p < 0.0001) with superior diagnostic performance.

Conclusions

The algorithm showed excellent performance in detecting COVID-19 pneumonia on chest CT images compared with resident radiologists.

Key Points

The higher sensitivity of deep learning model in detecting COVID-19 pneumonia were found compared with radiological residents on a per-lobe and per-patient basis.
The deep learning model improves diagnosis efficiency by shortening processing time.
The deep learning model can automatically calculate the volume of the lesions and whole lung.
Anhänge
Nur mit Berechtigung zugänglich
Literatur
1.
Zurück zum Zitat Kickbusch I, Leung G (2020) Response to the emerging novel coronavirus outbreak. BMJ 368:m406CrossRef Kickbusch I, Leung G (2020) Response to the emerging novel coronavirus outbreak. BMJ 368:m406CrossRef
3.
Zurück zum Zitat Zhu N, Zhang D, Wang W et al (2020) A nNovel cCoronavirus from pPatients with pPneumonia in China, 2019. N Engl J Med 382:727–733CrossRefPubMedCentral Zhu N, Zhang D, Wang W et al (2020) A nNovel cCoronavirus from pPatients with pPneumonia in China, 2019. N Engl J Med 382:727–733CrossRefPubMedCentral
5.
Zurück zum Zitat World Health Organization (2020) Coronavirus disease 2019 (COVID-19): situation report-39. https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200303-sitrep-43-ncov.pdf. Published (March 03, 2020) [Epub ahead of print] World Health Organization (2020) Coronavirus disease 2019 (COVID-19): situation report-39. https://​www.​who.​int/​docs/​default-source/​coronaviruse/​situation-reports/​20200303-sitrep-43-ncov.​pdf.​ Published (March 03, 2020) [Epub ahead of print]
6.
Zurück zum Zitat Mahase E (2020) Coronavirus covid-19 has killed more people than SARS and MERS combined, despite lower case fatality rate. BMJ 368:m641CrossRef Mahase E (2020) Coronavirus covid-19 has killed more people than SARS and MERS combined, despite lower case fatality rate. BMJ 368:m641CrossRef
11.
Zurück zum Zitat Long JB, Ehrenfeld JM (2020) The rRole of aAugmented iIntelligence (AI) in dDetecting and pPreventing the sSpread of nNovel cCoronavirus. J Med Syst 44:59CrossRefPubMedCentral Long JB, Ehrenfeld JM (2020) The rRole of aAugmented iIntelligence (AI) in dDetecting and pPreventing the sSpread of nNovel cCoronavirus. J Med Syst 44:59CrossRefPubMedCentral
12.
Zurück zum Zitat Health LD (2020) COVID-19 and artificial intelligence: protecting health-care workers and curbing the spread. Lancet Digital Health. S2589-7500(20):30054–30056 Health LD (2020) COVID-19 and artificial intelligence: protecting health-care workers and curbing the spread. Lancet Digital Health. S2589-7500(20):30054–30056
13.
Zurück zum Zitat Topol EJ (2019) High-performance medicine: the convergence of human and artificial intelligence. Nat Med 25:44–56CrossRef Topol EJ (2019) High-performance medicine: the convergence of human and artificial intelligence. Nat Med 25:44–56CrossRef
14.
Zurück zum Zitat Tomasev N, Glorot X, Rae JW et al (2019) A clinically applicable approach to continuous prediction of future acute kidney injury. Nature 572:116–119CrossRefPubMedCentral Tomasev N, Glorot X, Rae JW et al (2019) A clinically applicable approach to continuous prediction of future acute kidney injury. Nature 572:116–119CrossRefPubMedCentral
16.
Zurück zum Zitat Yu Q, Wang Y, Huang S et al (2020) Multicenter cohort study demonstrates more consolidation in upper lungs on initial CT increases the risk of adverse clinical outcome in COVID-19 patients. Theranostics 10(12):5641–5648CrossRefPubMedCentral Yu Q, Wang Y, Huang S et al (2020) Multicenter cohort study demonstrates more consolidation in upper lungs on initial CT increases the risk of adverse clinical outcome in COVID-19 patients. Theranostics 10(12):5641–5648CrossRefPubMedCentral
18.
Zurück zum Zitat Li Z, Zhang S, Zhang J, Huang K, Wang Y, Yu Y (2019) MVP-net: mMulti-view FPN with position-aware attention for deep universal lesion detection. International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) Li Z, Zhang S, Zhang J, Huang K, Wang Y, Yu Y (2019) MVP-net: mMulti-view FPN with position-aware attention for deep universal lesion detection. International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI)
19.
Zurück zum Zitat Ronneberger O, Fischer P, Brox T (2015) U-net: cConvolutional networks for biomedical image segmentation. In International Conference on Medical Image Computing and Computer-Assisted Intervention. Ronneberger O, Fischer P, Brox T (2015) U-net: cConvolutional networks for biomedical image segmentation. In International Conference on Medical Image Computing and Computer-Assisted Intervention.
20.
Zurück zum Zitat Wang XQ, Zhang QY, Zhou Z et al (2020) Evaluating multi-class segmentation errors with anatomical prior. IEEE International Symposium on Biomedical Imaging. Wang XQ, Zhang QY, Zhou Z et al (2020) Evaluating multi-class segmentation errors with anatomical prior. IEEE International Symposium on Biomedical Imaging.
23.
Zurück zum Zitat Ardila D, Kiraly AP, Bharadwaj S et al (2019a2019) End-to-end lung cancer screening with three-dimensional deep learning on low-dose chest computed tomography. Nat Med 25:954–-961 Ardila D, Kiraly AP, Bharadwaj S et al (2019a2019) End-to-end lung cancer screening with three-dimensional deep learning on low-dose chest computed tomography. Nat Med 25:954–-961
24.
Zurück zum Zitat Nam JG, Park S, Hwang EJ et al (2019) Development and vValidation of dDeep lLearning-based aAutomatic dDetection aAlgorithm for mMalignant pPulmonary nNodules on cChest rRadiographs. Radiology 290:218–228CrossRef Nam JG, Park S, Hwang EJ et al (2019) Development and vValidation of dDeep lLearning-based aAutomatic dDetection aAlgorithm for mMalignant pPulmonary nNodules on cChest rRadiographs. Radiology 290:218–228CrossRef
25.
Zurück zum Zitat Ardila D, Kiraly AP, Bharadwaj S et al (2019b2019) End-to-end lung cancer screening with three-dimensional deep learning on low-dose chest computed tomography. Nat Med 25:954–961 Ardila D, Kiraly AP, Bharadwaj S et al (2019b2019) End-to-end lung cancer screening with three-dimensional deep learning on low-dose chest computed tomography. Nat Med 25:954–961
26.
Zurück zum Zitat Rajpurkar P, Irvin J, Ball RL et al (2018) Deep learning for chest radiograph diagnosis: aA retrospective comparison of the CheXNeXt algorithm to practicing radiologists. PLoS Med 15:e1002686CrossRefPubMedCentral Rajpurkar P, Irvin J, Ball RL et al (2018) Deep learning for chest radiograph diagnosis: aA retrospective comparison of the CheXNeXt algorithm to practicing radiologists. PLoS Med 15:e1002686CrossRefPubMedCentral
27.
Zurück zum Zitat Lakhani P, Sundaram B (2017) Deep learning at chest radiography: automated classification of pulmonary tuberculosis by using convolutional neural networks. Radiology 284:574–582 Lakhani P, Sundaram B (2017) Deep learning at chest radiography: automated classification of pulmonary tuberculosis by using convolutional neural networks. Radiology 284:574–582
Metadaten
Titel
A deep learning approach to characterize 2019 coronavirus disease (COVID-19) pneumonia in chest CT images
verfasst von
Qianqian Ni
Zhi Yuan Sun
Li Qi
Wen Chen
Yi Yang
Li Wang
Xinyuan Zhang
Liu Yang
Yi Fang
Zijian Xing
Zhen Zhou
Yizhou Yu
Guang Ming Lu
Long Jiang Zhang
Publikationsdatum
02.07.2020
Verlag
Springer Berlin Heidelberg
Erschienen in
European Radiology / Ausgabe 12/2020
Print ISSN: 0938-7994
Elektronische ISSN: 1432-1084
DOI
https://doi.org/10.1007/s00330-020-07044-9

Weitere Artikel der Ausgabe 12/2020

European Radiology 12/2020 Zur Ausgabe

Update Radiologie

Bestellen Sie unseren Fach-Newsletter und bleiben Sie gut informiert.