Adenoma, Colonic Polyp, Colorectal Cancer, Colorectal Neoplasms
Conditions
Keywords
Artificial Intelligence, Colonoscopy, Colorectal Cancer Screening, Adenoma Detection Rate, Deep Learning
Brief summary
This study is a multi-center, randomized controlled trial designed to evaluate whether an artificial intelligence (AI) system can assist endoscopists to improve the detection rate of colorectal adenomas and cancers during colonoscopy compared to standard colonoscopy. Early screening and diagnosis are key to reducing the burden of colorectal cancer, but current colonoscopy has limitations, including the risk of missed lesions. This trial aims to determine if AI can enhance screening quality and diagnostic accuracy.
Detailed description
Colorectal cancer (CRC) screening is crucial for early detection and reducing mortality, yet current colonoscopy techniques face challenges such as variable adenoma detection rates (ADR) and the risk of missed diagnoses for subtle lesions. This study is a prospective, multi-center, parallel-group, randomized controlled trial aiming to validate the clinical value of an AI-assisted diagnostic system in improving screening quality. A total of 3300 participants will be randomized in a 1:1 ratio to undergo either AI-assisted colonoscopy (Experimental Group) or conventional high-definition colonoscopy (Control Group). The primary objective is to compare the ADR between the two groups. Secondary objectives include assessing the detection rate of advanced or specific types of polyps, the mean number of adenomas per procedure, and the impact of the AI system on both patient and physician satisfaction. The study will provide high-quality evidence for the standardized application of AI technology in CRC screening, with the ultimate goal of reducing the incidence and mortality of colorectal cancer.
Interventions
DEVICEAI-Assisted Colonoscopy
High-definition colonoscopy procedure with a real-time video analyzed artificial intelligence system.
Sponsors
Zhejiang University
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
SCREENING
Masking
SINGLE (Subject)
Masking description
Participants are blinded to group assignment. Endoscopists performing the procedures cannot be blinded, which is an inherent limitation of computer-aided detection trials. Histopathological assessment, Boston Bowel Preparation Scale scoring by a second observer, outcome adjudication, and statistical analysis are performed by blinded personnel; the statistician remains blinded until database lock.
Eligibility
Sex/Gender
ALL
Age
40 Years to 74 Years
Healthy volunteers
Yes
Inclusion criteria
1. Age 40 to 74 years, inclusive; men and women. 2. Scheduled for elective screening, diagnostic, or surveillance colonoscopy. 3. In generally good condition, mentally competent, and willing to provide written informed consent.
Exclusion criteria
1. Known contraindication to colonoscopy or to biopsy. 2. History of colon cancer or previous colonic surgery. 3. Bowel obstruction; confirmed or highly suspected colorectal cancer; inflammatory bowel disease; or active colorectal bleeding. 4. Colorectal polyposis, including a history of familial adenomatous polyposis. 5. Pregnancy or breastfeeding. 6. Participation in another drug or medical device clinical trial within 1 month before enrollment. 7. Any other condition that, in the investigator's judgment, makes the participant unsuitable for this trial. Withdrawal Criteria: 1. Withdrawal of consent by the participant. 2. Inadequate bowel preparation, defined as a Boston Bowel Preparation Scale (BBPS) total score below 6 or any segment score below 2. 3. Failure to reach the cecum. 4. Newly diagnosed active colitis during the procedure.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Adenoma Detection Rate (ADR) | From the day of the procedure up to 14 days post-procedure. | The proportion of participants with at least one histologically confirmed colorectal adenoma or adenocarcinoma. Detection and specimen collection occur during the colonoscopy, with final confirmation based on pathology reports. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Mean Adenomas Per Colonoscopy (APC) | From the day of the procedure up to 14 days post-procedure. | The mean number of histologically confirmed adenomas detected per participant. The number of polyps is counted during the procedure, but confirmation of adenoma status depends on pathology. |
| Advanced adenoma detection rate | From the day of the procedure up to 14 days post-procedure. | The proportion of participants with at least one histologically confirmed advanced adenoma, defined as an adenoma at least 10 mm in diameter, or with high-grade dysplasia, or with a villous component. |
| Serrated lesion detection rate | From the day of the colonoscopy up to 14 days after the procedure. | The proportion of participants with at least one histologically confirmed flat serrated lesion, including sessile serrated lesions with or without dysplasia and traditional serrated adenomas. Hyperplastic polyps are reported separately. |
| Patient Satisfaction Score | Assessed within 1 hour after completion of the colonoscopy procedure. | Assessed using a 5-point Likert scale questionnaire evaluating examination comfort and acceptance of the AI system (1=very unsatisfied, 5=very satisfied). |
| Proximal adenoma detection rate | From the day of the colonoscopy up to 14 days after the procedure. | The proportion of participants with at least one histologically confirmed adenoma located proximal to the splenic flexure. |
| Physician Satisfaction Score | Assessed within 1 hour after completion of the colonoscopy procedure. | Assessed using a 5-point Likert scale questionnaire evaluating the AI system's usability and reliability (1=very unsatisfied, 5=very satisfied). |
| Incidence of Procedure-Related Adverse Events | From the start of the procedure up to 30 days post-procedure. | The number and type of adverse events (AEs) and serious adverse events (SAEs), including but not limited to perforation, significant bleeding, and infection, are recorded. |
Countries
China
Contacts
CONTACTKefeng Ding, M.D., Ph.D.
STUDY_CHAIRKefeng Ding, M.D., Ph.D.
Second Affiliated Hospital, Zhejiang University, School of Medicine
Outcome results
None listed