A116 OPTIMIZING CROHN’S DISEASE ENDOSCOPIC SCORING OPERATING CHARACTERISTICS TO ASSIST WITH THE ADVENT OF ARTIFICIAL INTELLIGENCE

ABSTRACT: Abstract

Background

To date, no studies have used Artificial Intelligence (AI) to produce standard endoscopic scores for Crohn’s Disease (CD). These scores are essential to clinical practice and have a major cost in clinical trials.

Purpose

We aimed to re-examine the components of the Simple Endoscopic Score for Crohn’s Disease (SES-CD) and Modified Multiplier (MM-SES-CD) to see if scoring of individual colonic segments could be combined while maintaining predictive value accuracy, thus simplifying scoring for AI.

Method

Data from 279 participants in the UNITI and EXTEND trials as well as an infliximab biosimilar trial (NCT02096861) were used in this retrospective study. The primary outcome was endoscopic remission (ER-1) of the colon, defined as absence of ulcer

SUBMITTER: Galts C 

PROVIDER: S-EPMC9991391 | biostudies-literature | 2023 Mar

REPOSITORIES: biostudies-literature