Skip to content
Back to Publications

AI-Assisted scoring of UC histology may improve reliability of metrics relating to histologic remission and normalization: results from a pilot study of AIM-HI UC

Study Background


- Histologic remission is a key metric of disease activity for patients with ulcerative colitis (UC) associated with improved long-term outcomes [1].

- The composition of the UC inflammatory microenvironment informs histologic scoring [e.g., Geboes Score, Robarts Histopathology Index (RHI), Nancy Histologic Index (NHI)].

- However, inter- and intra-pathologist variability complicates this process, highlighting opportunities for more standardized, reproducible histologic scoring in UC.

- Here, we aimed to assess the impact of AI-assistance on agreement of pathologists with expert consensus scores using AIM-HI UC [2]*, an AI-powered algorithm that yields slide-level Geboes subgrade.

Conference

UEG Week

 

Authors

  • Shamshoian et al.