The Role of Artificial Intelligence Combined With Digital Cholangioscopy for Indeterminant and Malignant Biliary Strictures: A Systematic Review and Meta-analysis.
Thomas R McCarty, Raj Shah, Ronan P Allencherril et al.
✦ AI-curated · Sources linked
In 30 seconds
This systematic review and meta-analysis evaluated the diagnostic performance of artificial intelligence (AI) combined with digital cholangioscopy for indeterminant and malignant biliary strictures. Analyzing five studies with 675 lesions, the pooled sensitivity was 95% and specificity was 88%, indicating high diagnostic accuracy (SROC) of 97%.
Key findings
- Pooled sensitivity of AI-assisted cholangioscopy was 95% (95% CI: 85-98).
- Pooled specificity was 88% (95% CI: 76-94).
- Diagnostic accuracy (SROC) was 97% (95% CI: 95-98).
- Sensitivity analysis of CNN studies showed similar results with 95% sensitivity.
Why it matters
The integration of AI in cholangioscopy may significantly enhance the diagnostic yield for biliary strictures, potentially leading to improved patient outcomes and more accurate treatment decisions.