Autonomous Artificial Intelligence Versus AI Assisted Human Optical Diagnosis of Colorectal Polyps
Trial Snapshot
- Phase
- Not Applicable
- Status
- Recruiting
- Sponsor
- Enrollment
- 540
- Locations
- 1
- Primary Endpoint
- Accuracy of optical diagnosis, for polyps 1-5mm, compared with an agreed upon CADx-assisted diagnosis
Study Overview
Brief Summary
Computer-aided image-enhanced endoscopy can predict the nature of colorectal polyps with over 90% accuracy. This technology uses artificial intelligence (AI) to analyze video recordings of polyps, learning to make diagnoses in real-time. This means that doctors can get immediate predictions about small polyps during the procedure, reducing the need for separate pathology exams and saving costs, ultimately improving patient care.
Human and AI interactions are complex and a framework to reap synergistic effects CADx systems when used by humans to harness optimal performance needs to be established. AI solutions in medicine are usually developed to be used as assistive devices, however, then they rely on humans to correct AI errors. Optical polyp diagnosis is a complex task. Non experts usually achieve diagnostic accuracy in 70-80%. CADx systems have a similar diagnostic accuracy when used autonomously. Clinical evaluation of CADx systems showed that CADx assisted OD performs equally to the operator performance when using non CADx assisted OD. To harness a benefit of clinical CADx implementation we would have to find a way that synergies between human and CADx come into play to eliminate cases in which CADx assisted and/ or human OD results in low diagnostic accuracy and also addresses the problem of serrated polyp recognition.
Detailed Description
Our study hypothesis is that for CADx implementation, instead of using the high/low confidence framework, identifying cases with suboptimal diagnostic accuracy could be facilitated through identifying cases in which CADx and endoscopist disagreed in their diagnosis. Eliminating such cases might separate out cases with low accuracy when using CADx assisted OD. Since endoscopists have a high sensitivity but low specificity for serrated polyp OD, this framework will also allow us to implement a strategy to adequately manage serrated polyps found in the cohort.
Study Design
- Study Type
- Interventional
- Allocation
- Na
- Intervention Model
- Single Group
- Primary Purpose
- Diagnostic
- Masking
- None (Participant)
Eligibility Criteria
- Ages
- 45 Years to 80 Years (Adult, Older Adult)
- Sex
- All
- Accepts Healthy Volunteers
- No
Inclusion Criteria
- •Indication for full colonoscopy.
Exclusion Criteria
- •Known inflammatory bowel disease
- •Active colitis
- •coagulopathy
- •familial polyposis syndrome
- •poor general health, defined as an American Society of Anesthesiologists class >3
- •emergency colonoscopy
Arms & Interventions
All participants
The endoscopist will make an optical diagnosis (OD) prediction for all small polyps (up to 10 mm) in white light (WL). Then, the endoscopist will make another OD prediction using image enhanced endoscopy (IEE) modes. After that, CADx will be activated in the IEE mode and a CADx prediction will be documented. Finally, after seeing the CADx prediction, the endoscopist will make a final prediction, which can agree or disagree with the autonomous CADx one. Polyps will be resected and sent to a pathology lab, where a pathologic diagnosis (blinded to the endoscopist's predictions) will be rendered.
Intervention: CADx (AI) system (Other)
Outcomes
Primary Outcomes
Accuracy of optical diagnosis, for polyps 1-5mm, compared with an agreed upon CADx-assisted diagnosis
Time Frame: up to 100 weeks
Accuracy of optical diagnosis, for polyps 1-5mm, compared with an agreed upon CADx-assisted diagnosis , when histopathology results are used as the reference
Secondary Outcomes
- Accuracy of optical diagnosis, for polyps 1-10mm, compared with an agreed upon CADx-assisted diagnosis(up to 100 weeks)
- Test characteristics, including recall, specificity, positive and negative predictive values (PPV/NPV), and particularly the NPV of rectosigmoid neoplastic polyps.(up to 100 weeks)
- Agreement of surveillance interval recommendations of AI-A and AI-H compared with the pathology-based recommendations(up to 100 weeks)
- Proportion of patients for whom an immediate surveillance recommendation can be directly provided for each approach, and how often histopathology-based polyp examination would have been avoided.(up to 100 weeks)
- Variability of OD (AI-A and AI-H) across participating endoscopists.(up to 100 weeks)
- Cost-effectiveness of OD ((AI-A and AI-H)(up to 100 weeks)
Investigators
Daniel Von Renteln
Principal Investigator
Centre hospitalier de l'Université de Montréal (CHUM)