Artificial Intelligence to Implement Cost-saving Strategies for Colonoscopy Screening Based on in Vivo Prediction of Polyp Histology (SAVE)

Artificial Intelligence to Implement Cost-saving Strategies for Colonoscopy Screening Based on in Vivo Prediction of Polyp Histology

Saving by Artificial Intelligence for Virtual Endoscopy Biopsy Artificial Intelligence to Implement Cost-saving Strategies for Colonoscopy Screening Based on in Vivo Prediction of Polyp Histology

This three parallel-arms, randomized, multicenter trial is aimed at investigating the value of AI-assisted optical biopsy for differentiating between neoplastic and non-neoplastic polyps which will lead to the implementation of cost-saving strategies in screening programs. A cost-effectiveness analyses with the use of modern trial emulation analyses of large observational and clinical trial datasets and real-cost data will be conducted. To improve personalized treatment with a novel colonoscopy CADx risk-prediction tool, the investigators will even develop a novel deep learning algorithm for the optical biopsy of the alternative pathway of colorectal cancer carcinogenesis, namely the serrated pathway and develop cost-effectiveness models of AI-assisted optical biopsy in colorectal cancer screening that provides reliable information to identify cancer risk regardless of physicians' skill.

Standard, high-definition colonoscopy with the use of CADe assistance (GI-GENIUS, Medtronic; CAD-EYE, Fujifilm; WISE VISION ,NEC). All detected polyps regardless of size and optical diagnosis will be resected and sent to pathology.

Standard, high-definition colonoscopy with the use of CADe/CADx assistance (GI-GENIUS, Medtronic; CAD-EYE, Fujifilm; WISE VISION ,NEC). All detected polyps regardless of size and optical diagnosis will be resected and sent to pathology.

Standard, high-definition colonoscopy with the use of CADe/CADx assistance (GI-GENIUS, Medtronic; CAD-EYE, Fujifilm; WISE VISION, NEC). Polyps will be left in situ if diminutive (≤5 mm) in size, located in the rectum or sigma and optically diagnosed by the endoscopist using the system to be hyperplastic with high confidence, otherwise resected and sent to pathology.

Polyps will be left in situ if diminutive (≤5 mm) in size, located in the rectum or sigma and optically diagnosed by the endoscopist using the system to be hyperplastic with high confidence, otherwise resected and sent to pathology.

Non-inferiority in the Adenoma Detection Rate, defined as the proportion of participants with at least one adenoma (per-patient analysis) in the three arms, when adopting a cost-saving leave-in-situ strategy for non-neoplastic rectosigmoid diminutive polyps.

Negative Predictive Value for colorectal neoplasia when adopting a leave-in-situ strategy for rectosigmoid diminutive polyps based on the use of Artificial Intelligence.

Concordance between post-polypectomy surveillance based on strategies based on optical diagnosis with Artificial Intelligence and those based on histology, and when adopting a leave-in-situ strategy for colorectal diminutive polyps based on the use of Artificial Intelligence

Change in the cost of polypectomy and histology in screening programs, when implementing strategies based on Artificial Intelligence-based optical biopsy, without changes in benefit related with detection of colorectal neoplasia

Inclusion Criteria: All >40 years-old patients undergoing colonoscopy for selected indications Exclusion Criteria: patients with personal history of CRC, or IBD patients affected with Lynch syndrome or Familiar Adenomatous Polyposis. patients with inadequate bowel preparation (defined as Boston Bowel Preparation Scale <2 in any colonic segment). patients with previous colonic resection. patients on antithrombotic therapy, precluding polyp resection. patients who were not able or refused to give informed written consent.