Truace
Subscribe

The Role of Artificial Intelligence Combined With Digital Cholangioscopy for Indeterminant and Malignant Biliary Strictures: A Systematic Review and Meta-analysis

Background Current endoscopic retrograde cholangiopancreatography (ERCP) and cholangioscopic-based diagnostic sampling for indeterminant biliary strictures remain suboptimal. Artificial intelligence (AI)-based algorithms by means of computer vision in machine learning have been applied to cholangioscopy in an effort to improve diagnostic yield. The aim of this study was to perform a systematic review and meta-analysis to evaluate the diagnostic performance of AI-based diagnostic performance of AI-associated chol…

By the Truvace desk1 min read1 sourcePeer-reviewedTRV-2026-0679· cite & verify
The Role of Artificial Intelligence Combined With Digital Cholangioscopy for Indeterminant and Malignant Biliary Strictures: A Systematic Review and Meta-analysis
"The Endoscopy Suite" by Sam Blackman is licensed under CC BY 2.0. To view a copy of this license, visit https://creativecommons.org/licenses/by/2.0/.

In brief

By August 2026, researchers published a systematic review and meta-analysis of AI combined with digital cholangioscopy for indeterminate and malignant biliary strictures. The analysis pooled five studies totaling 675 lesions and 2,685,674 images, finding pooled sensitivity of 95%, specificity of 88%, and SROC accuracy of 97% for AI-assisted diagnosis.

Accurate diagnosis of biliary strictures is clinically important because current ERCP and cholangioscopic sampling remain suboptimal, and improved computer-vision support could reduce missed malignancies. The findings remain preliminary given the small number of studies and lack of reported prospective clinical outcomes, implementation data, or harms.

Main points

  1. Systematic review and meta-analysis included 5 studies covering 675 lesions and 2,685,674 cholangioscopic images.
  2. Four of five studies used deep learning with a convoluted neural network processing 30 to 60 frames per second.
  3. Bivariate model was used to compute pooled sensitivity, specificity, likelihood ratio, diagnostic odds ratio, and SROC curve.
  4. Sensitivity analysis of CNN-only studies (538 patients) showed 95% sensitivity, 88% specificity, and 97% SROC accuracy.

The gain

AI-based machine learning applied to cholangioscopy images achieved high pooled diagnostic performance for indeterminate and malignant biliary strictures, with 95% sensitivity and 88% specificity.

The rundown

The review searched per PRISMA and MOOSE guidelines and applied Cochrane Diagnostic Test Accuracy methodology with a bivariate model to pool results across studies.

Most systems analyzed were CNN-based deep learning models operating at 30 to 60 frames per second, evaluated on a combined dataset of over 2.6 million cholangioscopic images from 675 lesions.

Sources

  1. Peer-reviewedJournal of Clinical Gastroenterology2026-08-06

The debate