# AnX cites FDA data on AI-assisted capsule image review times

> After a separate FDA clearance for another software platform on September 25, 2026, the company highlighted earlier FDA-filed clinical results for ProScan.

- Canonical URL: https://newsverum.com/en/news/anx-proscan-fda-reading-times
- Publisher: Newsverum
- Section: Technology
- Country: United States
- Published: 2026-10-11T20:35:53+00:00
- Updated: 2026-10-11T20:35:53+00:00
- Event / data period: 2026-10-11
- Other editions: [Türkçe](https://newsverum.com/tr/news/anx-proscan-fda-reading-times.md), [العربية](https://newsverum.com/ar/news/anx-proscan-fda-reading-times.md), [Français](https://newsverum.com/fr/news/anx-proscan-fda-reading-times.md), [Español](https://newsverum.com/es/news/anx-proscan-fda-reading-times.md), [Deutsch](https://newsverum.com/de/news/anx-proscan-fda-reading-times.md), [Português](https://newsverum.com/pt/news/anx-proscan-fda-reading-times.md)

Texas-based capsule-endoscopy company AnX Robotica highlighted clinical results in the FDA file for NaviCam ProScan on October 11. ProScan received its De Novo decision on December 12, 2023, so the announcement was not a new clearance. AnX revisited the earlier ProScan data after the FDA separately cleared CapsoVision’s CapsoView and CapsoCloud software through a 510(k) decision on September 25, 2026. The ARTIC analysis in the FDA file included 133 patients; mean small-bowel image review time was 33 minutes 42 seconds with standard physician reading and 3 minutes 50 seconds when physicians used AI assistance. The reduction calculated from those figures is 88.6%.

The FDA file reported diagnostic yields for significant lesions of 62.4% with standard reading and 73.7% with AI-assisted physician reading. AnX’s release also presented timing results from the separate FDA review of the competing software, but the studies used different systems and designs and were not head-to-head tests; they do not establish direct superiority between products. The FDA says ProScan assists physician review, is not a stand-alone diagnostic tool and does not replace clinical decision-making.

## Data

### ProScan ARTIC study results

Participant count and review-time results from the prospective ARTIC study in the FDA De Novo file.

Methodology: Patient count and mean review times are from the FDA De Novo summary. The 88.6% reduction is calculated from the reported means of 33 minutes 42 seconds and 3 minutes 50 seconds, then rounded to one decimal place. The study ran from February 2021 through January 2022 at seven European centers; it is not a head-to-head comparison with the separate study of a competing product.

| label | value | unit | period | basis | period_start |
|---|---|---|---|---|---|
| Patients included in the ARTIC analysis | 133 | patients | 2022-01 | observed | 2021-02 |
| Reduction in small-bowel image review time with AI-assisted physician reading | 88.6 | percent | 2022-01 | observed | 2021-02 |

Sources: <https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN230027.pdf>, <https://www.prnewswire.com/news-releases/anx-robotica-highlights-documented-proscan-ai-workflow-performance-from-submission-to-fda-302905102.html> · Download data: [JSON](https://newsverum.com/api/v1/articles/anx-proscan-fda-reading-times/data/proscan-artic-study-results?lang=en&format=json) · [CSV](https://newsverum.com/api/v1/articles/anx-proscan-fda-reading-times/data/proscan-artic-study-results?lang=en&format=csv)

## Sources

- [AnX Robotica via PR Newswire](<https://www.prnewswire.com/news-releases/anx-robotica-highlights-documented-proscan-ai-workflow-performance-from-submission-to-fda-302905102.html>) (Source checked: 2026-10-11; Published: 2026-10-11)
- [U.S. Food and Drug Administration, De Novo Classification Request DEN230027](<https://www.accessdata.fda.gov/cdrh_docs/reviews/DEN230027.pdf>) (Source checked: 2026-10-11; Published: 2023-12-12)
- [U.S. Food and Drug Administration, 510(k) Summary K254230](<https://www.accessdata.fda.gov/cdrh_docs/pdf25/K254230.pdf>) (Source checked: 2026-10-11; Published: 2026-09-25)

---

Cite the canonical URL above. [Editorial standards and corrections](https://newsverum.com/en/editorial)