Background and Aims Histological disease activity in inflammatory bowel disease [IBD] is associated with clinical outcomes and is an important endpoint in drug development. We developed deep learning models for automating histological assessments in IBD. Methods Histology images of intestinal mucosa from phase 2 and phase 3 clinical trials in Crohn’s disease [CD] and ulcerative colitis [UC] were used to train artificial intelligence [AI] models to predict the Global Histology Activity Score [GHAS] for CD and Geboes histopathology score for UC. Three AI methods were compared. AI models were evaluated on held-back testing sets, and model predictions were compared against an expert central reader and five independent pathologists. Results The model based on multiple instance learning and the attention mechanism [SA-AbMILP] demonstrated the best performance among competing models. AI-modelled GHAS and Geboes subgrades matched central readings with moderate to substantial agreement, with accuracies ranging from 65% to 89%. Furthermore, the model was able to distinguish the presence and absence of pathology across four selected histological features, with accuracies for colon in both CD and UC ranging from 87% to 94% and for CD ileum ranging from 76% to 83%. For both CD and UC and across anatomical compartments [ileum and colon] in CD, comparable accuracies against central readings were found between the model-assigned scores and scores by an independent set of pathologists. Conclusions Deep learning models based upon GHAS and Geboes scoring systems were effective at distinguishing between the presence and absence of IBD microscopic disease activity.

BACKGROUND & AIMS: Endoscopic assessment of ulcerative colitis (UC) typically reports only the maximum severity observed. Computer vision methods may better quantify mucosal injury detail, which varies among patients. METHODS: Endoscopic video from the UNIFI clinical trial (A Study to Evaluate the Safety and Efficacy of Ustekinumab Induction and Maintenance Therapy in Participants With Moderately to Severely Active Ulcerative Colitis) comparing ustekinumab and placebo for UC were processed in a computer vision analysis that spatially mapped Mayo Endoscopic Score (MES) to generate the Cumulative Disease Score (CDS). CDS was compared with the MES for differentiating ustekinumab vs placebo treatment response and agreement with symptomatic remission at week 44. Statistical power, effect, and estimated sample sizes for detecting endoscopic differences between treat-ments were calculated using both CDS and MES measures. Endoscopic video from a separate phase 2 clinical trial replication cohort was performed for validation of CDS performance. RESULTS: Among 748 induction and 348 maintenance patients, CDS was lower in ustekinumab vs placebo users at week 8 (141.9 vs 184.3; P < .0001) and week 44 (78.2 vs 151.5; P < .0001). CDS was correlated with the MES (P < .0001) and all clinical com-ponents of the partial Mayo score (P < .0001). Stratification by pretreatment CDS revealed ustekinumab was more effective than placebo (P < .0001) with increasing effect in severe vs mild disease (-85.0 vs -55.4; P < .0001). Compared with the MES, CDS was more sensitive to change, requiring 50% fewer participants to demonstrate endoscopic differences between ustekinumab and placebo (Hedges' g 1/4 0.743 vs 0.460). CDS performance in the JAK-UC replication cohort was similar to UNIFI. CONCLUSIONS: As an automated and quantitative measure of global endoscopic disease severity, the CDS offers artificial intelligence enhancement of traditional MES capability to better evaluate UC in clinical trials and potentially practice.