Artificial Intelligence-assisted Common Bile Duct Stent Selection in Endoscopic Retrograde Cholangiopancreatography

Common bile duct stenosis is an important indication for endoscopic retrograde cholangiopancreatography(ERCP). Appropriate selection of bile duct stent size is not only conducive to successful stent implantation but also to improve the prognosis of patients. Currently, the selection of stent specifications is based on the operator's empirical estimation, which is not only not accurate but also increases the radiation exposure time, causing unnecessary harm to both the operator and the patient. Our objective is to develop an artificial intelligence algorithm to automatically select appropriate stent.

Study Overview

• Status: Recruiting
• Conditions: Endoscopic Retrograde Cholangiopancreatography; Stenosis of Bile Duct

Detailed Description

Endoscopic Retrograde Cholangiopancreatography (ERCP) is an operation with high risk. Common bile duct stone and stenosis are important indications. The quality control of ERCP is the key to improve its success rate and reduce complications, which has received great attention. In 2015, the American Society of Gastrointestinal Endoscopy/American College of Gastroenterology (ASGE/ACG) issued ERCP quality control indicators, among which biliary stent placement and radiographic fluoroscopy time are important intraoperative quality control indicators.

The selection of appropriate biliary stent size is not only conducive to successful stent implantation but also to improve the prognosis of patients. Choose a stent of appropriate length. The proximal side of the stent should be 1cm above the obstruction segment, and the distal tail should be located just outside the nipple. The length of the stent can be determined by measuring the distance between the proximal end of the obstruction and the nipple under X-ray.

Current stent size selection is based on the operator's empirical estimation :(1) estimate the distance by endoscope diameter or cone length or catheter marking; (2) By retracting the guidewire, calculate the distance of the guidewire retracting between two points to estimate the length of the stent.The long radiation exposure time results in unnecessary injuries to both the operator and the patient.

• Study Type: Observational
• Enrollment (Anticipated): 600

Contacts and Locations

• Study Contact: Name: Rui Ji, MD, PhD; Phone Number: 18560086103; Email: qljirui@email.sdu.edu.cn
• Study Contact Backup: Name: Yanqing Li, MD, PhD; Phone Number: 053182169385; Email: liyanqing@sdu.edu.cn

Study Locations

• China
  • Shandong
    • Jinan, Shandong, China, 250012
      • Recruiting
      • Department of Gastroenterology,QiLu Hospital,Shandong University
      • Contact: Rui Ji, MD, PhD; Phone Number: 18560086103; Email: qljirui@email.sdu.edu.cn

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: 18 years and older (Adult, Older Adult)
• Accepts Healthy Volunteers: No
• Genders Eligible for Study: All

• Sampling Method: Non-Probability Sample

Study Population

The eligible images was randomly devided into two groups. One group for training the algorithm and the other group for testing the algorithm.

Description

Inclusion Criteria:

• Patients older than 18 years old who underwent ERCP

Exclusion Criteria:

• failed cholangiopancreatography caused by failed intubation, gastric retention, duodenal disease and so on
• patients proved no stenosis in common bile duct
• poor cholangiograms due to the lack of contrast agent or insufficient filling of contrast agent (cholangiograms without the completed CBD or thumbnails)

Study Plan

How is the study designed?

Number of groups / cohorts

2

Cohorts and Interventions

Group / Cohort
group for training the algorithm; This group of images is used for training the algorithm of the artificial intelligence
group for testing the algorithm; This group of images is used for testing the algorithm of the artificial intelligence

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
The accuracy of the calculated length of the stents by the artificial intelligence	The length of the stent was calculated "the length from the stenosis to the papilla+2cm".The length of the stent selected by experts is the gold standard	6 months

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
The accuracy of the segmentation of the artificial intelligence	The accuracy of the segmentation of the common bile duct, duodenoscopy and stenosis lesions by the artificial intelligence	4 months

Collaborators and Investigators

• Sponsor: Shandong University

Publications and helpful links

General Publications

• Adler DG, Lieb JG 2nd, Cohen J, Pike IM, Park WG, Rizk MK, Sawhney MS, Scheiman JM, Shaheen NJ, Sherman S, Wani S. Quality indicators for ERCP. Gastrointest Endosc. 2015 Jan;81(1):54-66. doi: 10.1016/j.gie.2014.07.056. Epub 2014 Dec 2. No abstract available. Erratum In: Gastrointest Endosc. 2015 Apr;81(4):1060.
• Huang L, Lu X, Huang X, Zou X, Wu L, Zhou Z, Wu D, Tang D, Chen D, Wan X, Zhu Z, Deng T, Shen L, Liu J, Zhu Y, Gong D, Chen D, Zhong Y, Liu F, Yu H. Intelligent difficulty scoring and assistance system for endoscopic extraction of common bile duct stones based on deep learning: multicenter study. Endoscopy. 2021 May;53(5):491-498. doi: 10.1055/a-1244-5698. Epub 2020 Nov 9.

Study record dates

Study Major Dates

• Study Start (Actual): April 20, 2022
• Primary Completion (Anticipated): August 1, 2022
• Study Completion (Anticipated): April 1, 2023

Study Registration Dates

• First Submitted: April 3, 2022
• First Submitted That Met QC Criteria: April 3, 2022
• First Posted (Actual): April 11, 2022

Study Record Updates

• Last Update Posted (Actual): July 6, 2022
• Last Update Submitted That Met QC Criteria: July 2, 2022
• Last Verified: May 1, 2022

More Information

Terms related to this study

• Other Study ID Numbers: 2021-SDU-QILU-090

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No