A major UK trial is testing whether computer-assisted detection (CADe) truly transforms colonoscopy outcomes in everyday practice. Interim results from the NAIAD study—spanning 29 hospitals and 124 endoscopists—show CADe boosts adenoma detection rates, especially for nonexpert endoscopists, lifting ADR from under 20% to around 35%. But the effect vanishes once the system is switched off, raising questions about overreliance and whether the technology builds lasting skill.
Experts caution that while CADe offers an immediate lift, its benefits may be more fragile than hoped—highlighting both the promise and the pitfalls of AI in endoscopy.