The Implementation of Computer-aided Detection in Training Improves the Quality of Future Colonoscopies

Jagiellonian University

Status: Completed

Conditions: Quality Indicators, Health Care; Colonoscopy Diagnostic Techniques and Procedures; Artificial Intelligence (AI)

Treatments: Other: AI-enhanced endoscopy training; Other: Conventional endoscopy training

Study type: Observational

Funder types: Other

Identifiers

NCT06623331

2024.000.367

Details and patient eligibility

About

Computer-aided detection (CADe) based on artificial intelligence (AI) may improve colonoscopy quality. An increasing number of young endoscopists are trained in an AI environment. However its impact on trainees' future outcomes remains unclear. The study aimed to evaluate the quality indicators of endoscopists trained in an AI environment compared to those trained conventionally.

Full description

Computer-aided detection (CADe) based on artificial intelligence (AI) may improve colonoscopy quality. An increasing number of young endoscopists are trained in an AI environment. However its impact on trainees' future outcomes remains unclear. The study aimed to evaluate the quality indicators of endoscopists trained in an AI environment compared to those trained conventionally. A study included 6,000 adult patients who underwent a colonoscopy for various reasons. The study retrospectively evaluated the first 1,000 procedures performed by six endoscopists after completing training relying entirely on endoscopists' detection skills without AI enhancement. Three of those young endoscopists were trained with CADe, and three without additional assistance. Quality indicators were assessed in both groups. The morphology of detected polyps was evaluated to determine the influence of AI-enhanced training on laterally spreading tumors (LST) detection rate.

Enrollment: 6,000 patients

Sex: All

Ages: 18+ years old

Volunteers: No Healthy Volunteers

Inclusion criteria

• adult participants who underwent a colonoscopy for various reasons performed by specific endoscopists that were assessed in terms of quality indicators

Exclusion criteria

• a history of bowel resection
• confirmed inflammatory bowel disease
• suspicion of polyps or cancer in other imaging tests
• suspicion of familial adenomatous polyposis

Trial design

6,000 participants in 2 patient groups

Group A

Description:

Colonoscopies performed by endoscopists trained in AI-enhanced environment. The quality indicators are measured after completing training without AI enhancement.

Treatment:

Other: AI-enhanced endoscopy training

Group B

Description:

Colonoscopies performed by endoscopists trained conventionally. The quality indicators are measured after completing training without AI enhancement.

Treatment:

Other: Conventional endoscopy training