Post written by Yasuharu Maeda, MD, PhD, and Takanori Kuroki, MD, from the Digestive Disease Center, Showa University Northern Yokohama Hospital, Yokohama, Kanagawa, Japan.

In the assessment of inflammation and prediction of outcomes in ulcerative colitis (UC), artificial intelligence (AI)—assisted image-enhanced endoscopy offers objective and more precise predictions.

We developed an AI-assisted narrow-band imaging system called EB-UC3. This system provides a 2-class diagnosis (“vascular healing” or “vascular active”) based on magnified narrow-band imaging. Our study showed that patients diagnosed as AI-based vascular healing had significantly fewer clinical relapses than those diagnosed as AI-based vascular active, indicating that AI-based vascular healing could be a novel treat to target (T2T) in managing UC.

Our study also revealed no significant correlation between AI output and histologic Geboes grade. Despite this, both methods identified patients at high risk of clinical relapse. This suggests that, although AI-based vascular healing does not replace histologic remission, it could serve as a novel T2T in managing UC.

Using AI for vascular evaluation offers several advantages. It eliminates the need for biopsy specimen collection and wait for pathology results, thus reducing pathological effort and associated costs.

In addition, real-time evaluation with AI-based vascular healing could enable immediate decision-making regarding treatment interventions. For example, patients with an AI-based vascular-active status might need increased medication or adherence checks, whereas those with a vascular-healing status might be candidates for reduced medication as part of a T2T strategy.

Graphical Abstract

Read the full article online.

The information presented in Endoscopedia reflects the opinions of the authors and does not represent the position of the American Society for Gastrointestinal Endoscopy (ASGE). ASGE expressly disclaims any warranties or guarantees, expressed or implied, and is not liable for damages of any kind in connection with the material, information, or procedures set forth.