Background and Aims
Accurately diagnosing malignant biliary strictures (MBSs) as benign or malignant remains
challenging. It has been suggested that direct visualization and interpretation of
cholangioscopy images provide greater accuracy for stricture classification than current
sampling techniques (ie, brush cytology and forceps biopsy sampling) using ERCP. We
aimed to develop a convolutional neural network (CNN) model capable of accurate stricture
classification and real-time evaluation based solely on cholangioscopy image analysis.
Methods
Consecutive patients with cholangioscopy examinations from 2012 to 2021 were reviewed.
A CNN was developed and tested using cholangioscopy images with direct expert annotations.
The CNN was then applied to a multicenter, reserved test set of cholangioscopy videos.
CNN performance was then directly compared with that of ERCP sampling techniques.
Occlusion block heatmap analyses were used to evaluate and rank cholangioscopy features
associated with MBSs.
Results
One hundred fifty-four patients with available cholangioscopy examinations were included
in the study. The final image database comprised 2,388,439 still images. The CNN demonstrated
good performance when tasked with mimicking expert annotations of high-quality malignant
images (area under the receiver-operating characteristic curve, .941). Overall accuracy
of CNN-based video analysis (.906) was significantly greater than that of brush cytology
(.625, P = .04) or forceps biopsy sampling (.609, P = .03). Occlusion block heatmap analysis demonstrated that the most frequent image
feature for an MBS was the presence of frond-like mucosa/papillary projections.
Conclusions
This study demonstrates that a CNN developed using cholangioscopy data alone has greater
accuracy for biliary stricture classification than traditional ERCP-based sampling
techniques.
Graphical abstract
Graphical Abstract
Abbreviations:
AI (artificial intelligence), AUROC (area under the receiver-operating characteristic curve), CCA (cholangiocarcinoma), CI (confidence interval), CNN (convolutional neural network), HQB (high quality-benign), HQM (high quality-malignant), HQS (high quality-suspicious), LQ (Low quality), MBS (malignant biliary strictures), ROC (receiver-operating characteristic), UNI (uninformative)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 accessOne-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 EndoscopyAlready a print subscriber? Claim online access
Already an online subscriber? Sign in
Register: Create an account
Institutional Access: Sign in to ScienceDirect
References
- Increasing mortality in the United States from cholangiocarcinoma: an analysis of the National Center for Health Statistics Database.BMC Gastroenterol. 2016; 16: 117
- The impact of changed strategies for patients with cholangiocarcinoma in this millenium.HPB Surg. 2015; 2015736049
- Treatment and survival of resected and unresected distal cholangiocarcinoma: a nationwide study.Acta Oncol. 2019; 58: 1048-1055
- Cholangiocarcinoma trends, incidence, and relative survival in Khon Kaen, Thailand from 1989 through 2013: a population-based cancer registry study.J Epidemiol. 2019; 29: 197-204
- Incidence of benign lesions in patients resected for suspicious hilar obstruction.Br J Surg. 2001; 88: 48-51
- Incidence of benign pathology in patients undergoing hepatic resection for suspected malignancy.Surgeon. 2003; 1: 32-38
- Asia-Pacific consensus recommendations on endoscopic tissue acquisition for biliary strictures [Review].Aliment Pharmacol Ther. 2018; 48: 138-151
- Comparative effectiveness of biliary brush cytology and intraductal biopsy for detection of malignant biliary strictures: a systematic review and meta-analysis.Gastrointest Endosc. 2015; 81: 168-176
- Impact of trimodality sampling on detection of malignant biliary strictures compared with patients with primary sclerosing cholangitis.Gastrointest Endosc. 2022; 95: 884-892
- Diagnostic yield of EUS-guided FNA for malignant biliary stricture: a systematic review and meta-analysis.Gastrointest Endosc. 2016; 83: 290-298
- Trans-peritoneal fine needle aspiration biopsy of hilar cholangiocarcinoma is associated with disease dissemination.HPB. 2011; 13: 356-360
- Prospective cytological assessment of gastrointestinal luminal fluid acquired during EUS: a potential source of false-positive FNA and needle tract seeding.Am J Gastroenterol. 2010; 105: 1311-1318
- Single-operator cholangioscopy and targeted biopsies in the diagnosis of indeterminate biliary strictures: a systematic review.Gastrointest Endosc. 2015; 82: 608-614
- Is single-operator peroral cholangioscopy a useful tool for the diagnosis of indeterminate biliary lesion? A systematic review and meta-analysis.Gastrointest Endosc. 2015; 82: 79-87
- Cholangioscopy biopsies improve detection of cholangiocarcinoma when combined with cytology and FISH, but not in patients with PSC.Dig Dis Sci. 2020; 65: 1471-1478
- Safety, diagnostic accuracy and therapeutic efficacy of digital single-operator cholangioscopy.United Eur Gastroenterol J. 2018; 6: 902-909
- Digital, single-operator cholangiopancreatoscopy in the diagnosis and management of pancreatobiliary disorders: a multicenter clinical experience (with video).Gastrointest Endosc. 2016; 84: 649-655
- Performance of a fully disposable, digital, single-operator cholangiopancreatoscope.Endoscopy. 2017; 49: 651-658
- Diagnostic and therapeutic single-operator cholangiopancreatoscopy with SpyGlassDS: results of a multicenter retrospective cohort study.Surg Endosc. 2018; 32: 3981-3988
- Diagnostic accuracy and interobserver agreement of digital single-operator cholangioscopy for indeterminate biliary strictures.Gastrointest Endosc. 2021; 94: 1059-1068
- Application of artificial intelligence using a convolutional neural network for detecting gastric cancer in endoscopic images.Gastric Cancer. 2018; 21: 653-660
- Artificial intelligence and capsule endoscopy: automatic detection of vascular lesions using a convolutional neural network.Ann Gastroenterol. 2021; 34: 820-828
- Utilisation of artificial intelligence for the development of an EUS-convolutional neural network model trained to enhance the diagnosis of autoimmune pancreatitis.Gut. 2021; 70: 1335-1344
- Application of artificial intelligence using a novel EUS-based convolutional neural network model to identify and distinguish benign and malignant hepatic masses.Gastrointest Endosc. 2021; 93: 1121-1130
- Clinical diagnosis and staging of cholangiocarcinoma.Nat Rev Gastroenterol Hepatol. 2011; 8: 512-522
- Deep residual learning for image recognition.ArXiv. Epub. 2015; (Dec 10. Available at https://arxiv.org/abs/1512.03385. Accessed December 2, 2022)
Deng J, Dong W, Socher R, et al. Imagenet: a large-scale hierarchical image database. 2009 IEEE Conference on Computer Vision and Pattern Recognition; 2009. p. 248-255.
- Deep learning with Keras.Birmingham, UK: Packt Publishing Ltd, 2017
Martin A, Agarwal A, Barham P, et al. TensorFlow: large-scale machine learning on heterogeneous distributed systems. ArXiv. Epub 2015 Mar 16. Available at: https://arxiv.org/abs/1603.04467. Accessed December 2, 2022.
- Scikit-learn: machine learning in Python.Journal of Machine Learning Research. 2011; 12: 2825-2830
- Artificial intelligence for automatic diagnosis of biliary stricture malignancy status in single-operator cholangioscopy: a pilot study.Gastrointest Endosc. 2022; 95: 339-348
- Diversity in machine learning.IEEE Access. 2019; 7: 64323-64350
- Digital single-operator cholangioscopy (DSOC) improves interobserver agreement (IOA) and accuracy for evaluation of indeterminate biliary strictures: the Monaco classification.J Clin Gastroenterol. 2022; 56: e94-e97
- Artificial intelligence in endoscopy.Dig Dis Sci. 2022; 67: 1553-1572
- Real-time automated diagnosis of precancerous lesions and early esophageal squamous cell carcinoma using a deep learning model (with videos).Gastrointest Endosc. 2020; 91: 41-51
- Klatskin-mimicking lesions.Diagnostics. 2021; : 11
- Cholangioscopy and cholangioscopic forceps biopsy in patients with indeterminate pancreaticobiliary pathology.Clin Gastroenterol Hepatol. 2006; 4: 219-225
- Single-operator cholangioscopy for the diagnosis of bile duct lymphoma: a case report and brief review of the literature.Endoscopy. 2013; 45: E359-E360
Article info
Publication history
Published online: August 22, 2022
Accepted:
August 13,
2022
Received:
June 2,
2022
Footnotes
DISCLOSURE: The following authors disclosed financial relationships: N. B. Marya: Consultant for Boston Scientific. B. T. Petersen: Consultant for Olympus America and Pentax; research support from Boston Scientific and Ambu. R. Law: Consultant for ConMed, Medtronic and Boston Scientific. A. Storm: Research support from Apollo Endosurgery, Boston Scientific, Endo-TAGSS, Endogenex, and Enterasense; consultant for Apollo Endosurgery, Erbe, GI Dynamics, and Olympus. B. K. Abu Dayyeh: Consultant for Endogenex, Endo-TAGSS, Metamodix, BFKW, USGI, Cairn Diagnostics, Aspire Bariatrics, and Boston Scientific; research support from USGI, Cairn Diagnostics, Aspire Bariatrics, Boston Scientific, Medtronic, Endogastric Solutions, Apollo Endosurgery, and Spatz Medical; speaker for Olympus, Johnson and Johnson, Medtronic, and Endogastric Solutions. V. Chandrasekhara: Consultant for Boston Scientific, Covidien, LP, and Interpace diagnostics; shareholder for Nevakar Corporation. All other authors disclosed no financial relationships.
DIVERSITY, EQUITY, AND INCLUSION: We worked to ensure gender balance in the recruitment of human subjects. We worked to ensure ethnic or other types of diversity in the recruitment of human subjects. One or more of the authors of this paper self-identifies as an under-represented gender minority in science. While citing references scientifically relevant for this work, we actively worked to promote gender balance in our reference list. The author list of this paper includes contributors from the location where the research was conducted who participated in the data collection, design, analysis, and/or interpretation of the work.
Identification
Copyright
© 2023 by the American Society for Gastrointestinal Endoscopy
