A study published in Gut analyzed the effectiveness of Shanghai Wision AI's artificial intelligence-based computer aided detection system, which found significantly more adenomas and polyps than standard colonoscopy alone.
Researchers studied 1,089 patients, with 536 undergoing a routine colonoscopy and 522 undergoing a computer-aided diagnosis colonoscopy. The endoscopist in the CAD group used a real-time automatic polyp detection system to assist during the colonoscopy. Physicians checked every polyp the CAD system detected.
What you should know:
1. Concerning polyps, endoscopists in the control group detected 0.51 polyps per colonoscopy, while endoscopists in the CAD group detected 0.97 polyps per colonoscopy. There was a 1.89-fold increase in the mean number of polyps detected between the two groups. The polyp detection rate of the control group was 0.29, while the polyp detection rate of the CAD group was 0.45.
2. Concerning adenomas, endoscopists in the control group detected 0.31 adenomas per colonoscopy, while endoscopists in the CAD group detected 0.53 adenomas per colonoscopy. There was a 1.72-fold increase in the mean number of adenomas detected between the two groups. The adenoma detection rate of the control group was 0.20, while the adenoma detection rate of the CAD group was 0.29.
3. The AI system had 39 false alarms, an average of 0.075 false alarms per colonoscopy, and all polyps detected by the endoscopist were also detected by the automatic system.
4. Excluding additional biopsy time, procedure times between the two groups were similar.
5. Researchers found adenoma detection rate between the two groups was statistically significant in patients with normal bowel prep, but not in patients with excellent bowel prep.
"The results of this study clearly demonstrate that a high-performance, real-time automatic polyp detection system based on a deep learning algorithm can significantly increase polyp and adenoma detection, especially those that are 5 mm or less. Given its high accuracy, fidelity and stability, the Wision AI system could enable improved detection of colon polyps and adenomas in clinical practice," said Tyler Berzin, MD, GI and endoscopy co-director at Boston-based Harvard Medical School.