EYE-SIGHT trial and Adenoma Detection: CGH | September 2026
Introduction:
Colonoscopy quality traditionally focuses on bowel preparation, withdrawal technique, and increasingly AI-assisted detection. The EYE-SIGHT trial tested a different concept: whether changing the endoscopist’s visual-search pattern using real-time eye-tracking feedback could improve adenoma detection.
Why was this study needed?
Adenomas may be missed despite adequate withdrawal time and bowel preparation.
Endoscopists naturally tend to focus toward the centre of the monitor.
Flat or subtle lesions may appear in peripheral visual fields.
Visual-search behavior is a potentially modifiable component of colonoscopy quality.
Key Findings:
400 patients were randomized to conventional withdrawal or real-time peripheral-gaze feedback.
Adenomas per colonoscopy increased from 0.95 to 1.34 with eye-tracking guidance.
ADR increased from 39.2% to 53.3%.
Polyp detection increased from 47.2% to 66.5%.
The improvement occurred without a significant increase in observation time.
The greatest benefit appeared in detecting smaller and flat adenomas.
Detection of advanced adenomas and sessile serrated lesions was not significantly improved.
The proprietary system and inability to blind endoscopists mean broader validation is required before routine adoption.
Bottom Line:
Colonoscopy quality may depend not only on how long we look, but where we look. Real-time peripheral-gaze guidance substantially improved ADR without prolonging observation time, introducing visual-search behavior as a potentially modifiable quality metric.
