Automated and real-time validation of gastroesophageal varices under esophagogastroduodenoscopy using a deep convolutional neural network: a multicenter retrospective study (with video)

  • Author Footnotes
    ∗ Drs Chen, Wang, and Xiao contributed equally to this article.
    Mingkai Chen
    Footnotes
    ∗ Drs Chen, Wang, and Xiao contributed equally to this article.
    Affiliations
    Department of Gastroenterology, Renmin Hospital of Wuhan University, Wuhan, China

    Key Laboratory of Hubei Province for Digestive System Disease, Renmin Hospital of Wuhan University, Wuhan, China

    Hubei Provincial Clinical Research Center for Digestive Disease Minimally Invasive Incision, Renmin Hospital of Wuhan University, Wuhan, China
    Search for articles by this author
  • Author Footnotes
    ∗ Drs Chen, Wang, and Xiao contributed equally to this article.
    Jing Wang
    Footnotes
    ∗ Drs Chen, Wang, and Xiao contributed equally to this article.
    Affiliations
    Department of Gastroenterology, Renmin Hospital of Wuhan University, Wuhan, China

    Key Laboratory of Hubei Province for Digestive System Disease, Renmin Hospital of Wuhan University, Wuhan, China

    Hubei Provincial Clinical Research Center for Digestive Disease Minimally Invasive Incision, Renmin Hospital of Wuhan University, Wuhan, China
    Search for articles by this author
  • Author Footnotes
    ∗ Drs Chen, Wang, and Xiao contributed equally to this article.
    Yong Xiao
    Footnotes
    ∗ Drs Chen, Wang, and Xiao contributed equally to this article.
    Affiliations
    Department of Gastroenterology, Renmin Hospital of Wuhan University, Wuhan, China

    Key Laboratory of Hubei Province for Digestive System Disease, Renmin Hospital of Wuhan University, Wuhan, China

    Hubei Provincial Clinical Research Center for Digestive Disease Minimally Invasive Incision, Renmin Hospital of Wuhan University, Wuhan, China
    Search for articles by this author
  • Lianlian Wu
    Affiliations
    Department of Gastroenterology, Renmin Hospital of Wuhan University, Wuhan, China

    Key Laboratory of Hubei Province for Digestive System Disease, Renmin Hospital of Wuhan University, Wuhan, China

    Hubei Provincial Clinical Research Center for Digestive Disease Minimally Invasive Incision, Renmin Hospital of Wuhan University, Wuhan, China
    Search for articles by this author
  • Shan Hu
    Affiliations
    School of Resources and Environmental Sciences of Wuhan University, Wuhan, China
    Search for articles by this author
  • Shi Chen
    Affiliations
    Department of Gastroenterology, Wuhan Puren Hospital, Wuhan, China
    Search for articles by this author
  • Guodong Yi
    Affiliations
    Department of Gastroenterology, the Central Hospital of Enshi Tujia and Miao Autonomous Prefecture, Enshi, China
    Search for articles by this author
  • Wei Hu
    Affiliations
    Wuhan No. 1 Hospital, Wuhan, China
    Search for articles by this author
  • Xianmu Xie
    Affiliations
    Jingzhou Second People's Hospital, Jingzhou, China
    Search for articles by this author
  • Yijie Zhu
    Affiliations
    Department of Gastroenterology, Renmin Hospital of Wuhan University, Wuhan, China

    Key Laboratory of Hubei Province for Digestive System Disease, Renmin Hospital of Wuhan University, Wuhan, China

    Hubei Provincial Clinical Research Center for Digestive Disease Minimally Invasive Incision, Renmin Hospital of Wuhan University, Wuhan, China
    Search for articles by this author
  • Yiyun Chen
    Affiliations
    School of Resources and Environmental Sciences of Wuhan University, Wuhan, China
    Search for articles by this author
  • Yanning Yang
    Correspondence
    Professor Yanning Yang, Department of Ophthalmology, Renmin Hospital of Wuhan University, 99 Zhangzhidong Road, Wuhan 430060, Hubei Province, China.
    Affiliations
    Department of Ophthalmology, Renmin Hospital of Wuhan University, Wuhan, China
    Search for articles by this author
  • Honggang Yu
    Correspondence
    Reprint requests: Professor Honggang Yu, Department of Gastroenterology, Renmin Hospital of Wuhan University, 99 Zhangzhidong Road, Wuhan 430060, Hubei Province, China.
    Affiliations
    Department of Gastroenterology, Renmin Hospital of Wuhan University, Wuhan, China

    Key Laboratory of Hubei Province for Digestive System Disease, Renmin Hospital of Wuhan University, Wuhan, China

    Hubei Provincial Clinical Research Center for Digestive Disease Minimally Invasive Incision, Renmin Hospital of Wuhan University, Wuhan, China
    Search for articles by this author
  • Author Footnotes
    ∗ Drs Chen, Wang, and Xiao contributed equally to this article.

      Background and Aims

      Rupture of gastroesophageal varices is the most common fatal adverse event of cirrhosis. EGD is considered the criterion standard for diagnosis and risk stratification of gastroesophageal variceal bleeding. The aim of this study was to train and validate a real-time deep convolutional neural network (DCNN) system, named ENDOANGEL, for diagnosing gastroesophageal varices and predicting the risk of rupture.

      Methods

      After training with 8566 images of endoscopic gastroesophageal varices from 3021 patients and 6152 images of normal esophagus/stomach from 3168 patients, ENDOANGEL was also tested with independent images and videos. It was also compared with endoscopists in several aspects.

      Results

      ENDOANGEL, in contrast with endoscopists, displayed higher accuracy of 97.00% and 92.00% in terms of detecting esophageal varices (EVs) and gastric varices (GVs) in an image contest (97.00% vs 93.94% , P < .01; 92.00% vs 84.43%, P < .05). It also surpassed endoscopists for red color signs of EVs and red spots of GVs (84.21% vs 73.45%, P < .01; 85.26% vs 77.52%, P < .05). Moreover, ENDOANGEL achieved comparable performance in the determination of size, form, color, and bleeding signs. ENDOANGEL also had good performance in making treatment suggestions. With regard to predicting risk factors in multicenter videos, ENDOANGEL showed great stability.

      Conclusions

      Our data suggest that DCNNs were precise in detecting both EVs and GVs and performed excellently in uncovering the endoscopic risk factors of gastroesophageal variceal bleeding. Thus, the application of DCNNs will assist endoscopists in evaluating gastroesophageal varices more objectively and precisely. (Clinical trial registration number: ChiCTR1900023970.)

      Abbreviations:

      AUC (area under the curve), DCNN (deep convolutional neutral network), EV (esophageal varices), GV (gastric varices), RC (red color sign), ROC (receiver operating characteristic), VH (varices hemorrhage)
      To read this article in full you will need to make a payment

      Purchase one-time access:

      Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online access
      One-time access price info
      • For academic or personal research use, select 'Academic and Personal'
      • For corporate R&D use, select 'Corporate R&D Professionals'

      Subscribe:

      Subscribe to Gastrointestinal Endoscopy
      Already a print subscriber? Claim online access
      Already an online subscriber? Sign in
      Institutional Access: Sign in to ScienceDirect

      References

        • Garcia-Tsao G.
        • Abraldes J.G.
        • Berzigotti A.
        • et al.
        Portal hypertensive bleeding in cirrhosis: risk stratification, diagnosis, and management: 2016 practice guidance by the American Association for the study of liver diseases.
        Hepatology. 2017; 65: 310-335
        • European Association for the Study of the Liver
        EASL Clinical Practice Guidelines for the management of patients with decompensated cirrhosis.
        J Hepatol. 2018; 69: 406-460
        • North Italian Endoscopic Club for the Study and Treatment of Esophageal Varices
        Prediction of the first variceal hemorrhage in patients with cirrhosis of the liver and esophageal varices. A prospective multicenter study.
        N Engl J Med. 1988; 319: 983-989
        • Bambha K.
        • Kim W.R.
        • Pedersen R.
        • et al.
        Predictors of early re-bleeding and mortality after acute variceal haemorrhage in patients with cirrhosis.
        Gut. 2008; 57: 814-820
        • de Franchis R.
        Expanding consensus in portal hypertension: report of the Baveno VI Consensus Workshop: stratifying risk and individualizing care for portal hypertension.
        J Hepatol. 2015; 63: 743-752
        • Kalambokis G.N.
        • Baltayiannis G.
        • Christou L.
        • et al.
        Red signs and not severity of cirrhosis should determine non-selective β-blocker treatment in Child–Pugh C cirrhosis with small varices: increased risk of hepatorenal syndrome and death beyond 6 months of propranolol use.
        Gut. 2016; 65: 1228-1230
        • Takuma Y.
        • Nouso K.
        • Morimoto Y.
        • et al.
        Prediction of oesophageal variceal bleeding by measuring spleen stiffness in patients with liver cirrhosis.
        Gut. 2016; 65: 354-355
        • Ichikawa S.
        • Okamura S.
        • Tujigami K.
        • et al.
        Quantitative analysis of red color sign in the endoscopic evaluation of esophageal varices.
        Endoscopy. 2001; 33: 747-753
        • Jensen D.M.
        Endoscopic screening for varices in cirrhosis: findings, implications, and outcomes.
        Gastroenterology. 2002; 122: 1620-1630
        • Beppu K.
        • Inokuchi K.
        • Koyanagi N.
        • et al.
        Prediction of variceal hemorrhage by esophageal endoscopy.
        Gastrointest Endosc. 1981; 27: 213-218
        • Tajiri T.
        • Yoshida H.
        • Obara K.
        • et al.
        General rules for recording endoscopic findings of esophagogastric varices (2nd ed.).
        Digest Endosc. 2010; 22: 1-9
        • Sarin S.K.
        • Lamba G.S.
        • Kumar M.
        • et al.
        Comparison of endoscopic ligation and propranolol for the primary prevention of variceal bleeding.
        N Engl J Med. 1999; 340: 988-993
        • Tripathi D.
        • Stanley A.J.
        • Hayes P.C.
        • et al.
        U.K. guidelines on the management of variceal haemorrhage in cirrhotic patients.
        Gut. 2015; 64: 1680-1704
        • LeCun Y.
        • Bengio Y.
        • Hinton G.
        Deep learning.
        Nature. 2015; 521: 436
        • Wainberg M.
        • Merico D.
        • Delong A.
        • et al.
        Deep learning in biomedicine.
        Nature Biotechnol. 2018; 36: 829-838
        • Xiao J.
        • Wang F.
        • Wong N.-K.
        • et al.
        Global liver disease burdens and research trends: analysis from a Chinese perspective.
        J Hepatol. 2019; 71: 212-221
        • Mori Y.
        • Kudo S.E.
        • Misawa M.
        • et al.
        Real-time use of artificial intelligence in identification of diminutive polyps during colonoscopy: a prospective study.
        Ann Intern Med. 2018; 169: 357-366
        • Mori Y.
        • Kudo S.E.
        • Misawa M.
        • et al.
        Performance of a deep learning model vs human reviewers in grading endoscopic disease severity of patients with ulcerative colitis.
        JAMA Network Open. 2019; 2: e193963
        • Chen D.
        • Wu L.
        • Li Y.
        • et al.
        Comparing blind spots of unsedated ultrafine, sedated, and unsedated conventional gastroscopy with and without artificial intelligence: a prospective, single-blind, 3-parallel-group, randomized, single-center trial.
        Gastrointest Endosc. 2020; 91: 332-339.e333
        • Wu L.
        • Zhou W.
        • Wan X.
        • et al.
        A deep neural network improves endoscopic detection of early gastric cancer without blind spots.
        Endoscopy. 2019; 51: 522-531
        • Wu L.
        • Zhang J.
        • Zhou W.
        • et al.
        Randomised controlled trial of WISENSE, a real-time quality improving system for monitoring blind spots during esophagogastroduodenoscopy.
        Gut. 2019; 68: 2161-2169
        • Zhou J.
        • Wu L.
        • Wan X.
        • et al.
        A novel artificial intelligence system for the assessment of bowel preparation (with video).
        Gastrointest Endosc. 2020; 91: 428-435.e422
        • Ding S.
        • Li L.
        • Li Z.
        • et al.
        Smart electronic gastroscope system using a cloud–edge collaborative framework.
        Futur Gener Comput Syst. 2019; 100: 395-407
        • Bilal M.
        • Abougergi M.S.
        • Tayyem O.
        • et al.
        Thirty-day readmission after esophageal variceal hemorrhage and its impact on outcomes in the United States.
        J Clin Gastroenterol. 2020; 54: 477-483
        • Abougergi M.S.
        • Travis A.C.
        • Saltzman J.R.
        The in-hospital mortality rate for upper GI hemorrhage has decreased over 2 decades in the United States: a nationwide analysis.
        Gastrointest Endosc. 2015; 81: 882-888.e881
        • Garcia-Tsao G.
        • Sanyal A.J.
        • Grace N.D.
        • et al.
        Prevention and management of gastroesophageal varices and variceal hemorrhage in cirrhosis.
        Am J Gastroenterol. 2007; 102: 2086-2102
        • Miyoshi H.
        • Matsumoto A.
        • Oka M.
        • et al.
        Efficacy of prophylactic sclerotherapy in patients with hepatocellular carcinoma and varices negative for the red color sign.
        Gastrointest Endosc. 1997; 45: 498-502
        • Paquet K.J.
        • Kalk J.F.
        • Klein C.P.
        • et al.
        Prophylactic sclerotherapy for esophageal varices in high-risk cirrhotic patients selected by endoscopic and hemodynamic criteria: a randomized, single-center controlled trial.
        Endoscopy. 1994; 26: 734-740
        • Hosking S.W.
        • Johnson A.G.
        Gastric varices: a proposed classification leading to management.
        Br J Surg. 1988; 75: 195-196
        • Kim T.
        • Shijo H.
        • Kokawa H.
        • et al.
        Risk factors for hemorrhage from gastric fundal varices.
        Hepatology. 1997; 25: 307-312