Colorectal polyps Diminutive colorectal polyps Colorectal neoplasms Colorectal cancer Artificial intelligence, Computer-aided diagnosis, Diminutive colorectal polyps, DeepGI
Conditions
Interventions
Patients undergoing colonoscopy at participating study centers with at least one diminutive colorectal polyp measuring 5 mm or less. All eligible polyps will be evaluated in real time using the DeepGI
Diagnostic
Patients with diminutive colorectal polyps
Sponsors
Piyapoom Pakvisal
Eligibility
Sex/Gender
All
Age
18 Years to No maximum
Inclusion criteria
Inclusion criteria: 1. Undergoing elective colonoscopy for colorectal cancer screening or surveillance 2. Adequate bowel preparation 3. At least one colorectal polyp detected during the procedure. 4. The polyp can be adequately visualized under both WLI and NBI. 5. The polyp is resected during the procedure and sent for histopathological examination. 6. Ability to provide written informed consent.
Exclusion criteria
Exclusion criteria: 1. History of inflammatory bowel disease or hereditary polyposis syndromes. 2. Failure of the AI system to generate a prediction with sufficient confidence. 3. Excessive colonic peristalsis that prevents the AI from producing stable or continuous outputs. 4. Polyp location in a concealed or poorly visualized area that precludes adequate assessment by the AI system. 5. Pseudopolyps 6. Subepithelial lesions 7. Colonic mass finding 8. Pregnant patients
Design outcomes
Primary
| Measure | Time frame |
|---|---|
| Diagnostic performance At the index colonoscopy Diagnostic performance of DeepGI in distinguishing adenoma vs hyperplastic diminutive polyps using histopathology as reference standard under White light imaging (WLI) or naroow band imaging (NBI) | — |
Secondary
| Measure | Time frame |
|---|---|
| Performance comparison of WLI and NBI at Image interpretation to availability of the final histopathologic diagnosis Comparison of Diagnostic accuracy sensitivity specificity positive predictive value negative predictive value and area under receiver operating characteristic curve,Performance by polyp size and Paris morphology at Image interpretation to availability of the final histopathologic diagnosis Diagnostic accuracy sensitivity specificity positive predictive value negative predictive value and area under receiver operating characteristic curve by polyp size and Paris morphology,Performance in rectosigmoid area at Image interpretation to availability of the final histopathologic diagnosis Diagnostic accuracy sensitivity specificity positive predictive value negative predictive value and area under receiver operating characteristic curve in rectosigmoid colon,Performance across participation sites at Image interpretation to availability of the final histopathologic diagnosis Diagnostic accuracy sensitivity specificity positive predictive value negative predictive value and area under receiver operating characteristic curve across participating hospitals,Benchmark with Preservation and Incorporation of Valuable Endoscopic Innovations (PIVI) and Simple Optical Diagnosis Accuracy (SODA) criteria at Image interpretation to availability of the final histopathologic diagnosis Evaluate if diagnostic performance meet PIVI and SODA performance criteria or not | — |
Countries
Thailand
Contacts
Public ContactPiyapoom Pakvisal
Division of Gastroenterology, Department of Internal Medicine, Faculty of Medicine, Chulalongkorn University, Bangkok
Outcome results
None listed