The goal of this clinical trial is to evaluate the diagnostic yield of CADe in a consecutive population undergoing colonoscopy. The main question it aims to answer is the Adenoma Detection Rate (ADR). Participants undergoing colonoscopy for follow-up in a screening setting will be randomized in a 1:1 ratio to either receive Computer-Aided Detection (CADe) colonoscopy or a conventional colonoscopy (CC). GI Genius is the AI software that will be used in the present trial and is intended to be used as an adjunct to colonic endoscopy procedures to help endoscopists to detect in real time mucosal lesions (such as polyps and adenomas, including those with flat (non-polypoid) morphology) during standard screening and surveillance endoscopic mucosal evaluations. It is not intended to replace histopathological sampling as a means of diagnosis.Researchers will compare the CADe group and the CC-group to see if CAD-e can increase the ADR significantly.

Even if colonoscopy is considered the reference standard for the detection of colonic neoplasia, polyps are still missed. In large administrative cohort or case-control studies, the risk of early post-colonoscopy cancer appeared to be independently predicted by a relatively low polyp/adenoma detection rate. The adenoma detection rate among different endoscopists has been shown to be strictly related with the risk of post-colonoscopy interval cancer. When considering the very high prevalence of advanced neoplasia in the FIT-positive enriched population, the risk of post-colonoscopy interval cancer due to a suboptimal quality of colonoscopy may be substantial. Available evidence justifies therefore the implementation of efforts aimed at improving adenoma detection rate, based on retraining interventions and on the adoption of innovative technologies, designed to enhance the accuracy of the endoscopic examination.Nowadays, Artificial intelligence (AI) is gaining increased attention and investigation, since it seems to improve the quality of medical diagnosis and treatment. In the field of gastrointestinal endoscopy, two potential roles of AI in colonoscopy have been examined so far: automated polyp detection (CADe) and automated polyp histology characterization (CADx). CADe can minimize the probability of missing a polyp during colonoscopy, thereby improving the adenoma detection rate (ADR) and potentially decreasing the incidence of interval cancer. GI Genius is the AI software that will be used in the present trial. The software is developed by Medtronic Inc. (Dublin, Ireland) and is intended to be used as an adjunct to colonic endoscopy procedures to help endoscopists to detect in real time mucosal lesions (such as polyps and adenomas, including those with flat (non-polypoid) morphology) during standard screening and surveillance endoscopic mucosal evaluations. It is not intended to replace histopathological sampling as a means of diagnosis.

The objective of this study was to compare the diagnostic yield obtained by using CADe colonoscopy to the yield obtained by the standard colonoscopy (SC). As the risk of progression is higher for large than for small adenomas the specific contribution of the new technique in reducing the miss rate of large neoplasms represents an important outcome to be assessed in the study. In addition, given the suggested association of a higher miss-rate of serrated and flat lesions with an increased risk of early post-colonoscopy CRC, the benefit of the new technique in reducing the miss rate of these lesions will be assessed.

Inclusion Criteria:

• Patients aged 50 to 74 undergoing colonoscopy examination following a prior colonsocopy were polyps were found (follow-up) performed in the context of a regional mass-screening program.

Exclusion Criteria:

• Patients unwilling or unable to give informed consent.
• Patients reporting use of anti-platelet agents or anticoagulants precluding removal of polyps.