Predicting relapse in ulcerative colitis with AI-assisted endoscopy

AI-assisted colonoscopy could predict long-term clinical relapse in patients with ulcerative colitis (UC) who were in clinical remission. Moreover, this novel tool was compatible with a wide range of colonoscopes.

Dr Yasuharu Maeda (Showa University Northern Yokohama Hospital, Japan; University College Cork, Ireland) and colleagues developed an AI-assisted, image-enhanced endoscopy method to assess vascular healing and predict clinical relapse in patients with UC who were in clinical remission [1]. “Although there is already an AI-assisted endoscopy tool for the prediction of histologic remission approved in Japan, this tool, called EndoBRAIN-UC, is only applicable for the Endocyto colonoscope,” commented Dr Maeda [1,2]. The novel tool can be adapted to a wide range of standard colonoscopes.

The prospective cohort study at hand used the novel AI-enhanced colonoscopy model to diagnose ‘vascular-healing’ and predict clinical relapse based on the outcomes; 33 patients were considered to be in the ‘healing’ group, whereas 67 patients were considered to be in the ‘active disease’ group [1].

At 12 months, the relapse rate was 3.0% in the ‘healing’ group and 23.9% in the ‘active’ group (P=0.01), indicating that AI-assisted colonoscopy could identify patients at high risk of relapse. Furthermore, Dr Maeda showed that AI-enhanced colonoscopy, categorising patients according to their ‘vascular-healing’ status, may add to the AUC value of complete endoscopic remission for the prediction of relapse (AUC 0.65 to AUC 0.70).

“The current, novel, AI-based endoscopy system for diagnosing ‘vascular-healing’ identified patients with UC at high risk of relapse,” decided Dr Maeda.

Meet the Trialist: Read our interview with Dr Maeda

1. Maeda Y, et al. A novel artificial intelligence-assisted image enhanced endoscopy assesses accurately “vascular-healing” and predicts long-term clinical relapse in patients with ulcerative colitis: a prospective cohort study. OP16, 19^th Congress of ECCO, 21–24 February 2024, Stockholm, Sweden.
2. Maeda Y, et al. Gastrointest Endosc. 2022;95(4):747-756.e2.

Copyright ©2024 Medicom Medical Publishers



Posted on 18 April 202418 April 2024