Home > Gastroenterology > ECCO 2023 > Pearls of the Posters > Novel AI tool assessing mucosal inflammation achieves high correlation with histopathologists

Novel AI tool assessing mucosal inflammation achieves high correlation with histopathologists

Presented by
Prof. Laurent Peyrin-Biroulet, Nancy University Hospital, France
ECCO 2023

An artificial intelligence (AI) tool using image processing and machine learning algorithms consistently and accurately assessed mucosal inflammation in histological images of patients with ulcerative colitis (UC). AI tools may help to save time and solve the problem of inter- and intra-observer variability.

Mucosal healing is an emerging treatment goal in the management of UC. One of the most widely used scores to evaluate this endpoint is the Nancy score, which allows the assessment of acute and chronic inflammatory disease activity in the mucosa. However, scoring histological images requires pathologists’ training, which might not be available, especially in non-academic institutions or smaller hospitals. In addition, the subjectivity of the pathologist in the assessment may have been eliminated. To test this, Prof. Laurent Peyrin-Biroulet (Nancy University Hospital, France) and his team assessed whether an AI tool using image processing and machine learning algorithms, which assigns a Nancy index value to histopathology slides, could help assess histological disease activity [1].

Eight global sites submitted 600 UC histological images, which were added to the 200 images that were used in a preliminary smaller study. To train the algorithm, 90% of the probes were used and the other 10% were used for testing.

The cell and tissue regions of each training image were manually assessed by 3 histopathologists and assigned a Nancy index. These results were used to further train the AI, allowing the AI tool to fully characterise histological images, identify tissue types, cell types, cell numbers and locations, and measure the Nancy Index for each image.

The average intra-class correlation was 92.1% among the histopathologists and 91.1% between the histopathologists and the AI tool in all stages of disease progression. An even higher consensus was achieved at the extremes of the Nancy index.

This study shows that the robustness of the AI tool was substantially improved by the addition of a larger number of tissue samples while maintaining accuracy.

  1. Peyrin-Biroulet L, et al. Deployment of an artificial intelligence tool for precision medicine in ulcerative colitis: Preliminary data from 8 globally distributed clinical sites. P777, ECCO 2023, 01–04 March, Copenhagen, Denmark.

Copyright ©2023 Medicom Medical Publishers

Posted on