Artificial Intelligence (AI), Colonoscopy Diagnostic Techniques and Procedures, Quality Indicators, Health Care
Conditions
Keywords
Quality Indicatiors, Colonoscopy, Artificial Intelligence (AI), Computer-aided Detection (CADe), Adenoma detection rate (ADR)
Brief summary
Colorectal cancer is the second most common malignancy in the countries of the European Union. Colonoscopy is the primary method for detecting and preventing the development of colorectal cancer is endoscopic examination. This study aims to evaluate the impact of artificial intelligence on the detection rate of polyps and early stages of colorectal cancer.
Detailed description
Colorectal cancer is the second most common malignancy in the countries of the European Union. The primary method for detecting and preventing the development of colorectal cancer is endoscopic examination-colonoscopy, during which precancerous lesions such as adenomas and serrated polyps can be removed. The effectiveness of colonoscopy depends on the adenoma detection rate, which varies among endoscopists and is influenced by their skills and experience. It has been proven that high-quality colonoscopy prevents the omission of colorectal cancer, which might develop in the future as so-called interval cancer. A breakthrough in machine learning in recent years has enabled the development of commercial artificial intelligence systems. These systems aim to improve the detection rates of precancerous polyps and, consequently, potentially reduce the risk of developing colorectal cancer. Artificial intelligence is also expected to help standardize performance across endoscopic procedures of varying quality, thereby contributing to a reduction in colorectal cancer incidence in the future. This study aims to evaluate the impact of artificial intelligence on the detection rate of polyps and early stages of colorectal cancer.
Interventions
DEVICEComputer-aided detection (CADe)
Endo-Aid CADe system is an AI-assisted computer-aided lesion detection application on ENDO-AID hardware. It uses a complex algorithm created via a neural network developed and taught by Olympus. With this new app, the sophisticated machine learning system can alert the endoscopist in real-time when a suspicious lesion appears on the screen. The image from the vision processor is transferred to the CADe device. The computer application recognizes the shape of the polyps and marks their place on the monitor screen.
Sponsors
Jagiellonian University
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
DIAGNOSTIC
Masking
NONE
Eligibility
Sex/Gender
ALL
Age
50 Years to 65 Years
Healthy volunteers
No
Inclusion criteria
* Consent to participate in the study, * Age between 50 and 65 years, * Scheduled outpatient colonoscopy.
Exclusion criteria
* Previous colonoscopy, * History of colorectal surgery, * Ongoing biological therapy for any indication, * Primary sclerosing cholangitis, * Familial polyposis syndrome, * Chronic diarrhea, * Ulcerative colitis, * Crohn's disease.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Adenoma detection rate (ADR) | During the colonoscopy examination | The percentage of colonoscopies when at least one histologically proven adenoma was found. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Utility of artificial intelligence for both novice and experienced endoscopists | During the colonoscopy examination | The difference in adenoma detection rates (ADR) achieved with and without AI in trainees and expert endoscopists. |
| Assessing the morphology of polyps detected during colonoscopy | During the colonoscopy examination | Assessment of the differences in polyps' morphology detected in both arms of the study. |
| Cost analysis of procedures performed with the use of artificial intelligence | Through study completion, an average of 6 months | The assessment of cost-efficiency of AI implementation, including the increased cost of pathological evaluation and additional surveillance examinations. |
Countries
Poland
Contacts
Primary ContactZofia Orzeszko, MD
Outcome results
None listed