Wision A.I. secures FDA nod for polyp-spotting colonoscopy tech

One down, one to go. After locking down its first FDA clearance, Wision A.I. is already lining up its next regulatory score.

The first authorization was granted to EndoScreener, an artificial intelligence software system that aims to identify potentially precancerous polyps during a colonoscopy. Next up, the Shanghai-based company said it’s preparing an FDA pre-submission for another AI-based tool that looks for colon polyps with high-grade dysplasia in tissue sample slides.

“Obtaining 510(k) clearance from the FDA for EndoScreener and initiating a U.S.-based investigational validation of the new tool for histopathological diagnosis demonstrates how our broadened, enriched AI portfolio can help to reduce colorectal cancer and improve patient outcomes,” said JingJia Liu, Wision’s co-founder and CEO.

Wision sat on the FDA news for several months, as EndoScreener was actually cleared in November, according to its regulatory filing. The authorization arrived barely a week after the technology also racked up European CE mark approval.

The FDA’s decision was based on a clinical study published last fall that showed how using the deep learning AI during a colonoscopy could result in fewer missed polyps—clumps of cells that build up on the lining of the colon that are typically benign but can be precursors to colon cancer. The technology was able to spot about 32% more polyps than manual observation alone, resulting in an adenoma miss rate of 20%, compared to more than 31% for standard methods.

The results are especially promising amid findings that up to one-third of precancerous lesions may go undetected during a colonoscopy, potentially delaying treatment for colorectal cancer, which is among the most common forms of cancer for both men and women worldwide.

Wision’s colonoscopy-enhancing technology received its FDA nod arrived shortly after the agency greenlit a similar system distributed by Medtronic in April 2021. Both of these also echo other software developed by Olympus, Iterative Scopes, Docbot and more.

Like EndoScreener, these systems also rely on AI to process images from a colonoscope in real time, highlighting polyps and other lesions along the walls of the colon throughout the procedure.

Also like EndoScreener, Medtronic’s GI Genius—which was originally developed by Cosmo Pharmaceuticals—has demonstrated significant improvements over manual colonoscopy analyses in clinical trials. Study data published earlier this month found that using the AI system resulted in a polyp miss rate of just 15%, slashing the AI-less method’s rate in half.