Abstract:Background and objectives: In spite of rapid growth of artificial intelligence (AI) in digestive endoscopy in lesion detection and characterization, the role of AI in inflammatory bowel disease (IBD) endoscopy is not clearly defined. We aimed at systematically reviewing the role of AI in IBD endoscopy and identifying future research areas. Methods: We searched the PubMed and Embase database using keywords ("artificial intelligence" OR "machine learning" OR "computer-aided" OR "convolutional neural network") AND ("inflammatory bowel disease" OR "ulcerative colitis" OR "Crohn's") AND ("endoscopy" or "colonoscopy" or "capsule endoscopy" or "device assisted enteroscopy") between 1975 and September 2023 and identified 62 original articles for detailed review. Review articles, consensus guidelines, case reports/series, editorials, letter to the editor, non-peer-reviewed pre-prints and conference abstracts were excluded. The quality of the included studies was assessed using the MI-CLAIM checklist. Results: The accuracy of AI models (25 studies) to assess ulcerative colitis (UC) endoscopic activity ranged between 86.54% and 94.5%. AI-assisted capsule endoscopy reading (12 studies) substantially reduced analyzable images and reading time with excellent accuracy (90.5% to 99.9%). AI-assisted analysis of colonoscopic images can help differentiate IBD from non-IBD, UC from non-UC and UC from Crohn's disease (CD) (three studies) with 72.1%, 98.3% and > 90% accuracy, respectively. AI models based on non-invasive clinical and radiologic parameters could predict endoscopic activity (three studies). AI-assisted virtual chromoendoscopy (four studies) could predict histologic remission and long-term outcomes. Computer-assisted detection (CADe) of dysplasia (two studies) is feasible along with AI-based differentiation of high from low-grade IBD neoplasia (79% accuracy). AI is effective in linking electronic medical record data (two studies) with colonoscopic videos to facilitate widespread machine learning. Conclusion: AI-assisted IBD endoscopy has the potential to impact clinical management by automated detection and characterization of endoscopic lesions. Large, multi-center, prospective studies and commercially available IBD-specific endoscopic AI algorithms are warranted.

Evolution and New Horizons of Endoscopy in Inflammatory Bowel Diseases

Technological advances in inflammatory bowel disease endoscopy and histology

Innovative Diagnostic Endoscopy in Inflammatory Bowel Diseases: From High-Definition to Molecular Endoscopy

Evolving role of endoscopy in inflammatory bowel disease: Going beyond diagnosis

Advanced Endoscopic Imaging Techniques for the Study of Colonic Mucosa in Patients with Inflammatory Bowel Disease

AI and endoscopy/histology in UC: the rise of machine

Role of artificial intelligence in imaging and endoscopy for the diagnosis, monitoring and prognostication of inflammatory bowel disease: a scoping review protocol

Artificial intelligence: A new tool in the pathologist's armamentarium for the diagnosis of IBD

Artificial intelligence and inflammatory bowel disease: Where are we going?

AI-luminating Artificial Intelligence in Inflammatory Bowel Diseases: A Narrative Review on the Role of AI in Endoscopy, Histology, and Imaging for IBD

Artificial intelligence in endoscopy related to inflammatory bowel disease: A systematic review

Advancements in Endoscopic Resection for Colitis-Associated Colorectal Neoplasia in Inflammatory Bowel Disease: Turning Visible into Resectable

From Staining Techniques to Artificial Intelligence: A Review of Colorectal Polyps Characterization

Imaging of Mucosal Inflammation: Current Technological Developments, Clinical Implications, and Future Perspectives

Image enhanced colonoscopy: updates and prospects-a review

Interventional inflammatory bowel disease: current and future practice

Therapeutic endoscopy: Recent updates and future directions

Automated Endoscopic Diagnosis in IBD: The Emerging Role of Artificial Intelligence

Scoring endoscopic disease activity in IBD: artificial intelligence sees more and better than we do

Inflammatory bowel disease genomics, transcriptomics, proteomics and metagenomics meet artificial intelligence

Opening the doors of precision medicine: novel tools to assess intestinal barrier in inflammatory bowel disease and colitis-associated neoplasia