Artificial intelligence-aided colonoscopy does not increase cancer detection

Adenoma and polyp detection rates (ADR/PDR) are lower when using artificial intelligence-assisted colonoscopy (AIAC), according to a study published in the November issue of The American Journal of Gastroenterology.

Idan Levy, MD, of the Sheba Medical Center at Ramat Gan and Tel Aviv University School of Medicine in Israel, and colleagues compared ADR/PDR over a six-month period before and after the introduction of AIAC in a large-volume center. Data were included for 2,175 pre-AIAC colonoscopies and 1,969 AIAC colonoscopies performed by 30 different gastroenterologists.

The researchers found that ADRs were lower in the AIAC group than in the pre-AIAC group (30.3 versus 35.2 percent); In addition, the AIAC group had a lower median number of adenomas per colonoscopy (0.60 versus 0.67).

PDR was also lower in the AIAC group than in the pre-AIAC group (36.5 versus 40.9 percent), and the number of polyps per colonoscopy was also lower (0.76 versus 0.83). Total procedure time was significantly shorter in the AIAC group (22 versus 24 minutes); for colonoscopies without any evidence of polyps, the median procedure time was also significantly shorter in the AIAC group (21 versus 22 minutes).

“Our results underscore the importance of examining AI-enabled technologies in the context of real-world use to uncover unanticipated impacts on the machine-human interface,” the authors write. “These results point to further challenges and directions for integrating AI-assisted colonoscopy into routine practice.”

Two authors disclosed financial ties to the biopharmaceutical industry.

Copyright © 2022 Health Day. All rights reserved.