Artificial Intelligence for Adenoma and Polyp Detection During Screening and Surveillance Colonoscopy: A...

Area of Science:

• Gastroenterology
• Oncology
• Medical Imaging

Background:

• Colorectal cancer (CRC) is a leading cause of cancer mortality in Kuwait.
• High adenoma detection rate (ADR) is crucial for colonoscopy's CRC prevention effectiveness.
• The role of computer-aided detection (CADe) in enhancing colonoscopy screening remains under investigation.

Purpose of the Study:

• To evaluate the effectiveness of a CADe system in improving ADR during screening colonoscopies.
• To assess secondary outcomes including polyp detection rate (PDR), adenoma per colonoscopy (APC), and polyp per colonoscopy (PPC).

Main Methods:

• A randomized-controlled trial (RCT) involving 102 patients aged 45+ undergoing screening or surveillance colonoscopy.
• Patients were randomized to either high-definition white-light (HD-WL) colonoscopy or HD-WL with a CADe system.
• Primary outcome was ADR; secondary outcomes included PDR, APC, PPC, and polyp characterization accuracy.

Main Results:

• The CADe group showed a non-significant trend towards higher ADR (47.1% vs. 37.3%, p=0.3).
• Polyp detection rate (PDR) and polyp per colonoscopy (PPC) were significantly higher in the CADe group (p=0.02 and p=0.04, respectively).
• No significant differences were observed in adenoma per colonoscopy (APC) or polyp characterization accuracy.

Conclusions:

• The AI-powered CADe system demonstrated a trend in improving ADR for colorectal cancer screening in Kuwait.
• CADe significantly enhanced the detection of diminutive polyps during colonoscopy.
• Larger, multicenter studies are recommended to confirm the impact of CADe on adenoma detection.