To conduct an single blinded, non-randomized, prospective, single center trial to validate the performance of a novel state-of-the-art Artificial Intelligence model (AI-Model) for colorectal lesion detection during routine diagnostic colonoscopy and to evaluate its feasibility in daily endoscopy. Consecutive patients referred for a screening, surveillance or diagnostic colonoscopy will be included

All procedures will be performed with high-definition endoscopes; Bowel preparation will be conducted according to usual local practices. During colonoscopy, the colonoscope will be first advanced to the cecum in all patients as confirmed by identification of the appendicular orifice and ileocecal valve or by intubation of the ileum, as per the standard of care by experienced endoscopists. During the insertion, no action will be taken. After cecal intubation is performed, the colonoscope will be slowly withdrawn to the splenic flexure by the primary endoscopists. Real time AI detection model will be activated with the output displayed in real time on a separate monitor and will be only viewed by an independent investigator, who is an experienced endoscopists (or a person trained in polyp recognition). The primary endoscopists will be blinded to the AI real time detection result

All detected polyps will be removed or with biopsy taken during this examination and sent for histopathology examination.

Inclusion Criteria:

1. Patients undergoing colonoscopy of age 18 to 80 years.
2. Patient who has read the informed consent form, has understood the relevant aspects of the study, and grants his/her authorization to participate by signing the informed consent form before the inclusion in the study and the performance of any procedure
3. Consecutive adult patients scheduled to undergo routine colonoscopy at AIG

Exclusion Criteria:

1. Females who are pregnant
2. Patients who are unsuitable for any other reason to participate in the study in the opinion of the investigator
3. Recent colonoscopy within past 12 month
4. History of inflammatory bowel disease
5. History of colorectal cancer
6. Previous bowel resection (apart from appendectomy)
7. Peutz-Jeghers syndrome, familial adenomatous polyposis or other polyposis syndromes
8. Bleeding tendency or severe comorbid illnesses for which polypectomy is considered unsafe.

Our pilot series showed AI assistance can detect up to 80 % of missed lesions with sample variance at around 0.032. We hypothesize that the adenoma/polyp miss rate of conventional colonoscopy can be reduced by 50% to 75% with AI assistance. Assuming 4% patients may be excluded. The sample size is estimated to be 381 patients in total with a power of 95% and a significance level of 0.05.

An interim analysis will be performed when the first 150 patients are recruited to verify the sample size estimation