研究數據來源於2020年1月1日至2020年3月18日之間，從洪湖和南昌的醫院回顧性收集了408例確診的COVID-19患者的初步CT圖像。神州醫療聯合南方醫院將洪湖市人民醫院的303例患者的數據作為訓練數據，將南昌大學第一附屬醫院的105例患者的數據作為測試數據。開發並驗證了使用多實例學習和殘差卷積神經網絡（ResNet34）的基於深度學習的模型。分別使用接收器工作特性曲線和混淆矩陣評估模型的判別能力和預測準確性。
圖1.本研究中介紹的多實例學習框架的概述。構建了從重新採樣的CT數據中裁剪出來的補丁以進行訓練，並通過最大合併所有屬於同一補丁的補丁來執行患者級別的預測。
基於深度學習的模型的曲線下面積（AUC）為0.987（95％置信區間[CI]：0.968–1.00），在訓練集中的準確性為97.4％，而AUC為0.892 （0.828–0.955），測試集中的準確度為81.9％。在入院時非嚴重COVID-19的患者的亞組分析中，該模型在洪湖和南昌亞組中分別獲得0.955（0.884–1.00）和0.923（0.864–0.983）的AUC，準確度分別為97.0和81.6％。
結論：我們基於深度學習的模型可以使用CT成像準確預測COVID-19患者的疾病嚴重程度和疾病進展，為指導臨床治療提供依據。
表1：洪湖和南昌人群的人口統計學和基線特徵
圖2.基於CT圖像的深度學習模型的培訓和驗證過程，用於疾病嚴重度預測任務。（A）交叉熵損失和（B）準確性針對100個訓練迭代作圖。交叉熵損失接近0.03，最終驗證準確性為87％。
圖3.接收機工作特性（ROC）曲線和混淆矩陣，用於預測訓練和測試集中的疾病嚴重程度。嚴重程度的預測結果通過ROC曲線顯示。（A）在訓練集中，多實例學習模型的曲線下面積（AUC）為0.987（CI：0.967-1.00）；（B）在測試集中，模型的AUC為0.892（CI：0.828-0.955）。混淆矩陣指示針對（C）訓練和（D）測試數據集的多實例學習模型分類的預測質量。
圖4.預測洪湖和南昌亞組疾病進展的受試者工作特徵（ROC）曲線和混淆矩陣。通過ROC曲線顯示疾病進展的預測結果。在洪湖和南昌隊列中入院時表現出非嚴重症狀的患者分別被分為洪湖和南昌亞組。（A）在洪湖亞組中，多實例學習模型的曲線下面積（AUC）為0.955（CI：0.884–1.00）；（B）在南昌亞組中，模型的AUC為0.923（CI：0.864–0.983）。混淆矩陣指示（C）訓練和（D）的多實例學習模型分類的預測質量 測試數據集。
建議確診COVID-19患者入院後應立即進行CT篩查
採用深度學習方法MIL，使用定量CT數據準確預測COVID-19的疾病嚴重程度。通過利用廉價且廣泛可用的測試，模型可用於識別疾病早期階段處於疾病進展高風險的患者，這對於進行早期幹預，預防疾病進展和降低死亡率具有重要的實際意義。因此，我們建議確診的COVID-19患者入院後應立即進行CT篩查，以便醫生可以使用我們的模型來確定發生嚴重疾病的風險。如果結果表明患者病情可能惡化，則應在疾病嚴重程度增加之前考慮更緊密的監測和早期幹預。
準確獲得重症患者的數據，探尋現象與數據間的聯繫，研究現象與數據背後蘊含的意義，是神州醫療進行醫療大數據工作最重要的基礎。COVID-19的蔓延讓神州醫療更加堅定了「深度挖掘數據價值 更好服務醫患群體」的信念。神州醫療將以高質量的大數據為基礎，繼續面向醫院、政府、藥企、保險公司等提供基於醫療大數據、雲平臺、藥企大數據、精準醫療、影像及AI的全方位、全周期的數位化解決方案，服務臨床科研、助力公共決策、推動人工智慧在醫療健康領域的進一步落地，為更好地實現健康中國戰略提供技術支撐。
參考資料
1. Anthimopoulos, M., Christodoulidis, S., Ebner, L., Christe, A., and Mougiakakou, S. (2016). Lung pattern classification for interstitial lung diseases using a deep convolutional neural network. IEEE Trans. Med. Imaging 35, 1207–1216. doi: 10.1109/tmi.2016.2535865
2. Campanella, G., Hanna, M. G., Geneslaw, L., Miraflor, A., Silva, V. W. K., Busam, K. J., et al. (2019). Clinical-grade computational pathology using weakly supervised deep learning on whole slide images. Nat. Med. 25, 1301–1309. doi: 10.1038/s41591-019-0508-1
3. Carpenter, J., and Bithell, J. (2000). Bootstrap confidence intervals: when, which, what? A practical guide for medical statisticians. Stat. Med. 19, 1141–1164.
4. Chang, Y.-C., Yu, C.-J., Chang, S.-C., Galvin, J. R., Liu, H.-M., Hsiao, C.-H., et al. (2005). Pulmonary sequelae in convalescent patients after severe acute respiratory syndrome: evaluation with thin-section CT. Radiology 236, 1067–1075. doi: 10.1148/radiol.2363040958
5. Chen, J., Wu, L., Zhang, J., Zhang, L., Gong, D., Zhao, Y., et al. (2020). Deep learning-based model for detecting 2019 novel coronavirus pneumonia on high-resolution computed tomography: a prospective study. medRxiv [Preprint], doi: 10.1101/2020.02.25.20021568
6. Chen, N., Zhou, M., Dong, X., Qu, J., Gong, F., Han, Y., et al. (2020). Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study. Lancet 395, 507–513. doi: 10.1016/s0140-6736(20)30211-7
7. Das, K. M., Lee, E. Y., Enani, M. A., AlJawder, S. E., Singh, R., Bashir, S., et al. (2015). CT correlation with outcomes in 15 patients with acute Middle East respiratory syndrome coronavirus. Am. J. Roentgenolog. 204, 736–742. doi: 10.2214/AJR.14.13671
8. Depeursinge, A., Chin, A. S., Leung, A. N., Terrone, D., Bristow, M., Rosen, G., et al. (2015). Automated classification of usual interstitial pneumonia using regional volumetric texture analysis in high-resolution computed tomography. Invest. Radiol. 50, 261–267. doi: 10.1097/rli.0000000000000127
9. Feng, Z., Yu, Q., Yao, S., Luo, L., Duan, J., Yan, Z., et al. (2020). Early prediction of disease progression in 2019 novel coronavirus pneumonia patients outside Wuhan with CT and clinical characteristics. medRxiv [Preprint], doi: 10.1101/2020.02.19.20025296
10. Fu, M., Yi, S.-L., Zeng, Y., Ye, F., Li, Y., Dong, X., et al. (2020). Deep learning-based recognizing COVID-19 and other common infectious diseases of the lung by chest CT scan images. medRxiv [Preprint], doi: 10.1101/2020.03.28.20046045
11. Gong, J., Ou, J., Qiu, X., Jie, Y., Chen, Y., Yuan, L., et al. (2020). A tool to early predict severe 2019-novel coronavirus pneumonia (COVID-19): a multicenter study using the risk nomogram in wuhan and guangdong, China. medRxiv [Preprint] doi: 10.1101/2020.03.17.20037515
12. Gupta, S., Agrawal, A., Gopalakrishnan, K., and Narayanan, P. (2015). 「Deep learning with limited numerical precision,」 in Proceedings of the 32nd International Conference on Machine Learning (ICML-15), Vienna. 1737–1746.
13. Huang, G., Liu, Z., Van Der Maaten, L., and Weinberger, K. Q. (2017). 「Densely connected convolutional networks,」 in Proceedings of the IEEE Computer Society Conference Computer Vision Pattern Recognition, Honolulu, HI.
14. Krizhevsky, A., Sutskever, I., and Hinton, G. E. (2012). 「Imagenet classification with deep convolutional neural networks,」 in In Advances in Neural Information Processing Systems, Sydney. 1097–1105.
15. Li, L., Qin, L., Xu, Z., Yin, Y., Wang, X., Kong, B., et al. (2020a). Artificial intelligence distinguishes COVID-19 from community acquired pneumonia on Chest CT. Radiology 2020:200905. doi: 10.1148/radiol.2020200905
16. Li, L., Yang, L., Gui, S., Pan, F., Ye, T., Liang, B., et al. (2020b). Association of clinical and radiographic findings with the outcomes of 93 patients with COVID-19 in Wuhan, China. Theranostics 10:6113. doi: 10.7150/thno.46569
17. Liang, W., Liang, H., Ou, L., Chen, B., Chen, A., Li, C., et al. (2020). Development and validation of a clinical risk score to predict the occurrence of critical illness in hospitalized patients with COVID-19. JAMA Intern. Med. 12:e202033. doi: 10.1001/jamainternmed.2020.2033
18. Mukherjee, H., Ghosh, S., Dhar, A., Obaidullah, S., Santosh, K., and Roy, K. (2020). Shallow convolutional neural network for COVID-19 outbreak screening using chest X-rays. TechRxiv [Preprint], doi: 10.36227/techrxiv.12156522
19. National Health Commission (2020). Diagnosis and treatment protocol for novel coronavirus pneumonia (Trial Version 7). Chin. Med. J. 133, 1087–1095. doi: 10.1097/CM9.0000000000000819
20. Rajinikanth, V., Dey, N., Raj, A. N. J., Hassanien, A. E., Santosh, K., and Raja, N. (2020). Harmony[-search and otsu based system for coronavirus disease (COVID-19) detection using lung CT scan images. arXiv [Preprint]. Available online at: https://arxiv.org/abs/2004.03431 (accessed April 6, 2020).
21. Robin, X., Turck, N., Hainard, A., Tiberti, N., Lisacek, F., Sanchez, J.-C., et al. (2011). pROC: an open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinform. 12:77. doi: 10.1186/1471-2105-12-77
22. Santosh, K. (2020). AI-driven tools for coronavirus outbreak: need of active learning and cross-population train/test models on multitudinal/multimodal data. J. Med. Syst. 44, 1–5. doi: 10.1007/s10916-020-01562-1
23. Schmidhuber, J. (2015). Deep learning in neural networks: an overview. Neural Netw. 61, 85–117. doi: 10.1016/j.neunet.2014.09.003
24. Simonyan, K., and Zisserman, A. (2014). Very deep convolutional networks for large-scale image recognition. arXiv [Preprint], Available online at: https://arxiv.org/abs/1409.1556 (accessed September 4, 2014).
25. Sohrabi, C., Alsafi, Z., O』Neill, N., Khan, M., Kerwan, A., Al-Jabir, A., et al. (2020). World Health Organization declares global emergency: a review of the 2019 novel coronavirus (COVID-19). Int. J. Surg. 76, 71–76. doi: 10.1016/j.ijsu.2020.02.034
26. Song, Y., Zheng, S., Li, L., Zhang, X., Zhang, X., Huang, Z., et al. (2020). Deep learning Enables Accurate Diagnosis of Novel Coronavirus (COVID-19) with CT images. medRxiv [Preprint], doi: 10.1101/2020.02.23.20026930
27. Wang, Y., Dong, C., Hu, Y., Li, C., Ren, Q., Zhang, X., et al. (2020). Temporal changes of CT findings in 90 patients with COVID-19 pneumonia: a longitudinal study. Radiology 2020:200843. doi: 10.1148/radiol.2020200843
28. World Health Organization (2020). Coronavirus Disease 2019 (COVID-19) Situation Report - 81. Geneva: WHO.
29. Xie, S., Girshick, R., Dollár, P., Tu, Z., and He, K. (2017). 「Aggregated residual transformations for deep neural networks,」 in Proceedings of the IEEE Comput Society Conference Computer Vision and Pattern Recognition, New York, NY.
30. Xu, P. P., Tian, R. H., Luo, S., Zu, Z. Y., Fan, B., Wang, X. M., et al. (2020). Risk factors for adverse clinical outcomes with COVID-19 in China: a multicenter, retrospective, observational study. Theranostics 10:6372. doi: 10.7150/thno.46833
31. Yang, Y., Lu, Q., Liu, M., Wang, Y., Zhang, A., Jalali, N., et al. (2020). Epidemiological and clinical features of the 2019 novel coronavirus outbreak in China. medRxiv [Preprint], doi: 10.1101/2020.02.10.20021675
32. Zheng, C. (2020). Time course of lung changes at chest CT during recovery from Coronavirus Disease 2019 (COVID-19). Radiology 295, 715–721. doi: 10.1148/radiol.2020200370
33. Zheng, C., Deng, X., Fu, Q., Zhou, Q., Feng, J., Ma, H., et al. (2020). Deep Learning-based Detection for COVID-19 from Chest CT using Weak Label. medRxiv [Preprint], doi: 10.1101/2020.03.12.20027185
34. Zhu, N., Zhang, D., Wang, W., Li, X., Yang, B., Song, J., et al. (2020). A novel coronavirus from patients with pneumonia in China, 2019. N. Engl. J. Med. 382, 727–733. doi: 10.1056/NEJMoa2001017
版權聲明：本文是神州醫療與南方醫科大學南方醫院合作的科研成果，如需轉載請後臺留言，經本平臺同意後註明轉載來源。
微信號 : DigitalHealthChina
掃碼關注