Artificial Intelligence, Colonic Adenoma, Colonic Neoplasms, Colonic Polyp, Colonoscopy
Conditions
Keywords
colo-rectal cancer screening, artificial intelligence assisted colonoscopy
Brief summary
Cancer can develop in the colon, or large bowel. Examination of the colon with a tube fitted with a camera is called a colonoscopy. Colonoscopy allows detection of small growths in the colon, called polyps. Polyps can often be removed during colonoscopy. Some of these polyps are called adenomas and can become cancer after several years. A good colonoscopy aims to find and take out as many of these polyps as possible. A quality indication of colonoscopy is the adenoma detection rate (ADR). It should be high, meaning many polyps are detected and taken out. New artificial intelligence devices to assist colonoscopy seem to increase the ADR, and maybe help prevent cancer even better than normal colonoscopy. The goal of this clinical trial is to compare the ADR when using standard colonoscopy to the ADR with artificial intelligence (AI)-assisted colonoscopy.
Detailed description
The colon is a part of the bowel where colon cancer can develop. It is possible to prevent colon cancer by doing a screening test called a colonoscopy. The colonoscopy procedure allows detection of polyps which can often be removed during the procedure. Some of these polyps are called adenomas and can become cancer after several years. A good colonoscopy aims to find and take out as many of these polyps as possible. A quality indication of colonoscopy is the adenoma detection rate (ADR). It should be high, meaning many polyps are detected and taken out. New artificial intelligence devices to assist colonoscopy seem to increase the ADR, and maybe help prevent cancer even better than normal colonoscopy. The goal of this clinical trial is to compare the ADR when using standard colonoscopy to the ADR with artificial intelligence (AI)-assisted colonoscopy. Patients who are scheduled to have screening colonoscopy and who agree to participate, and are aged 45 years or more, will be randomly assigned to receive either standard colonoscopy or AI-assisted colonoscopy. The main objective of this study is the difference in the ADR between a conventional colonoscopy procedure and an AI-assisted colonoscopy procedure. Secondary objectives will compare the 2 groups (conventional colonoscopy and AI-assisted colonoscopy) regarding polyp size, polyp quantity, polyp histology (laboratory analysis of the polyp that was taken out), polyp dysplasia grade (how close the polyp is to cancer), polyp location in the colon, endoscopist experience (older or younger doctor), time of day and colonoscopy preparation quality (how clean the bowel is). With these results we can show that AI-assisted colonoscopy is useful or not to help better prevent colon cancer.
Interventions
OTHERconventional colonoscopy procedure
Study subjects in this interventional arm will undergo conventional colonoscopy.
OTHERartificial intelligence-assisted colonoscopy procedure
Study subjects in this interventional arm will undergo colonoscopy done with a commercially-available module that uses artificial intelligence to highlight suspected polyps on the screen during colonoscopy. This module also attempts to characterize the detected polyp as adenomatous or not. The detection and characterization of polyps is in real time, during the procedure.
Sponsors
Chirec
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
DIAGNOSTIC
Masking
NONE
Intervention model description
Arm 1- Conventional colonoscopy Arm 2- Colonoscopy assisted by artificial intelligence system
Eligibility
Sex/Gender
ALL
Age
45 Years to 74 Years
Healthy volunteers
Yes
Inclusion criteria
* Patient (woman or man) candidate for a screening colonoscopy - Age: 45 to 74 years included * Absence of inflammatory bowel disease * Absence of significant digestive symptoms indicating colonoscopy (i.e. screening is the only indication for the examination) * Patient able to understand the concept of the study and agreeing to participate
Exclusion criteria
* Patient outside the inclusion age * All
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Adenoma Detection Rate in Conventional versus Artificial Intelligence-Assisted Colonoscopy | 1 day | The main objective of this study is the difference in the detection rate of colorectal adenomas in individuals over 45 years old during colonoscopies for colon cancer screening between a conventional colonoscopy procedure (CCP) and a colonoscopy procedure with AI (ACP). |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| The difference in the detection rate of colorectal adenomas according to the number per group (n=1-2/n= 3-10/n >10) between a conventional colonoscopic procedure (CCP) and a colonoscopy procedure with AI (ACP). | 1 day | The number of adenomas detected during the procedure will be compared between a conventional colonoscopic procedure (CCP) and a colonoscopy procedure with AI (ACP). |
| The difference in the detection rate of colorectal adenomas based on histology by group (hyperplastic/conventional adenomas/serrated adenomas/adenocarcinoma) between a conventional colonoscopic procedure (PCC) and a colonoscopy procedure with AI. | 1 month | Adenomas have different histologic types. After obtaining histology, the detection rate of the various histologic types will be compared between a conventional colonoscopic procedure (CCP) and a colonoscopy procedure with AI (ACP). |
| The difference in detection rate of colorectal adenomas based on dysplastic grade by group | 1 day | The difference in detection rate of colorectal adenomas based on dysplastic grade by group (conventional or serrated adenomas/low-grade dysplasia/high-grade dysplasia/adenocarcinoma) or Narrow Band Imaging International Colorectal Endoscopic (NICE) (stages 1, 2 or 3) classification or Japan Narrow Band Imaging Expert Team (JNET) classification (stages 1, 2A and 2B, or 3) between a conventional colonoscopy procedure (CCP) and a colonoscopy procedure with AI (ACP). |
| The difference in the detection rate of colorectal adenomas according to size by group (5 mm/6-9 mm/>10 mm) between a conventional colonoscopic procedure (CCP) and a colonoscopy procedure with AI (ACP). | 1 day | Detection rate of diminutive, small, and large polyps will be compared between a conventional colonoscopic procedure (CCP) and a colonoscopy procedure with AI (ACP). |
| The difference in the detection rate of colorectal adenomas according to experience by group of colonoscopists | 1 day | The difference in the detection rate of colorectal adenomas according to experience by group (Colonoscopists aged 45-55 years/56-65 years/\>66 years) between a conventional colonoscopic procedure (CCP) and a conventional colonoscopic procedure (CCP) colonoscopy with AI (ACP). |
| The difference in the detection rate of colorectal adenomas depending on the time of day . | 1 day | The difference in the detection rate of colorectal adenomas depending on the time of day (before 11:00/between 11:00 and 15:00/after 15:00) and the number of colonoscopies (Colonoscopy from 1 to 6 per half-day) per group (hyperplastic/conventional adenomas/scalloped adenomas/adenocarcinoma) between a conventional colonoscopic procedure (PCC) and a colonoscopy procedure with AI (ACP). |
| The difference in the detection rate of colorectal adenomas according to colonic preparation by group (Boston Score 9/6-8/<6) between a conventional colonoscopic procedure (PCC) and a colonoscopy procedure with AI (ACP). | 1 day | The difference in the adenoma detection rate according to the quality of colonic preparation will be compared between a conventional colonoscopic procedure (CCP) and a colonoscopy procedure with AI (ACP). |
| The difference in the detection rate of colorectal adenomas depending on the location by group (rectum/left colon/transverse colon/right colon) between a conventional colonoscopic procedure (CCP) and a colonoscopy procedure with AI (ACP). | 1 day | Location of adenomas will be compared between a conventional colonoscopic procedure (CCP) and a colonoscopy procedure with AI (ACP). |
Countries
Belgium
Contacts
Primary ContactErik Francois, M.D.
Backup ContactPhilippe Langlet, M.D.
Outcome results
None listed