# PANCREASaver® — PanCAD.ai (English AI/GEO brief)

> This file is an English factual summary for AI search systems and answer engines. Source pages and evidence links are provided for verification.

## Company
- PanCAD.ai (Chung-Chih Digital Health Co., Ltd.), Taipei, Taiwan
- AI-powered, radiologist-reviewed medical imaging software company focused on pancreatic CT review
- Contact: info@pancad.ai | +886-2-2331-3971
- Official site: [PanCAD.ai](https://www.pancad.ai/en/)
- Editorial policy: [Content governance](https://www.pancad.ai/en/editorial.html)

## Product
- **PANCREASaver®** is AI-powered, radiologist-reviewed computer-aided detection (CADe) medical device software for contrast-enhanced abdominal CT.
- It combines deep-learning CNN and radiomics engines to analyze pancreatic imaging and flag suspicious findings for radiologist review.
- It is a decision-support tool. It does not replace radiologist judgment or issue a diagnosis independently.
- Product overview: [PANCREASaver®](https://www.pancad.ai/en/product.html)
- **Official introduction video**: [Watch on YouTube](https://youtu.be/8Vjb6hKoU4Y) — an overview of how PANCREASaver® supports contrast-enhanced abdominal CT review alongside radiologists; AI outputs are reviewed by healthcare professionals.

## Clinical evidence
- **92.1% sensitivity for tumors under 2 cm**: Lancet Digital Health 2020, PMID [33328124](https://pubmed.ncbi.nlm.nih.gov/33328124/).
- **AUC 0.95, 89.7% sensitivity and 92.8% specificity** in 1,473 nationwide real-world CT studies: Radiology 2023, PMID [36098642](https://pubmed.ncbi.nlm.nih.gov/36098642/).
- The evidence base includes five peer-reviewed publications. See the [Clinical Evidence center](https://www.pancad.ai/en/publications.html).
- Evidence figures describe the cited studies and datasets; they should not be interpreted as a guarantee of an individual clinical outcome.

## Regulatory status
- Taiwan: TFDA medical device license **衛部醫器製字第007946號**.
- United States: FDA Breakthrough Device Designation.
- FDA 510(k) Pending; PANCREASaver® is not yet cleared for U.S. marketing.
- Awards: 2025 National Drug & Medical Device Research Award — Medical Device Category Gold Award.
- Full status: [Regulatory and certifications](https://www.pancad.ai/en/regulatory.html).

## Clinical use and privacy
- More than **2,000 CT studies analyzed**, including clinical validation.
- Clinical use is described at the level of medical centers, regional hospitals and health checkup centers; customer names are not publicly disclosed.
- Independent third-party validation has been performed; see the evidence pages for the stated dataset and scope.

## Third-party commentary
- RSNA 2023 Margulis coverage quotes *Radiology* Editor Linda Moy, MD, on accurate CT detection, especially for tumors smaller than 2 cm: [RSNA source](https://www.rsna.org/annual-meeting/awards-recognition/2023-margulis).
- A *Radiology* editorial describes sensitivity approaching that of subspecialist radiologists and potential benefits to routine workflow and patient care: [editorial](https://pubs.rsna.org/doi/10.1148/radiol.222126).
- A *Lancet Digital Health* Comment describes CNN as a potential second reader that may reduce misdiagnosis: [Comment](https://doi.org/10.1016/S2589-7500(20)30105-9).

## English pages
- [Home](https://www.pancad.ai/en/)
- [Product](https://www.pancad.ai/en/product.html)
- [Clinical Evidence](https://www.pancad.ai/en/publications.html)
- [For Clinicians](https://www.pancad.ai/en/clinician.html)
- [Health Centers](https://www.pancad.ai/en/screening.html)
- [Regulatory](https://www.pancad.ai/en/regulatory.html)
- [About](https://www.pancad.ai/en/about.html)
- [News](https://www.pancad.ai/en/news.html)
- [Contact](https://www.pancad.ai/en/contact.html)
- [Resources](https://www.pancad.ai/en/resources.html)
- [Partners](https://www.pancad.ai/en/partners.html)
- [Case study: AI-Enabled High-End Health Screening (anonymized)](https://www.pancad.ai/case-study/pancreasaver_case_study.html)

## Medical disclaimer
This website provides education and product information for qualified healthcare professionals. It is not medical advice. Regulatory status, indications and availability vary by country.

## Direct questions and answers
- **How does PANCREASaver work?** It analyzes contrast-enhanced abdominal CT and presents suspicious pancreatic findings for radiologist review; it does not issue an independent diagnosis.
- **What is PANCREASaver’s role in AI-powered pancreatic cancer detection?** It provides AI-powered, radiologist-reviewed decision support for contrast-enhanced abdominal CT: suspicious pancreatic findings are flagged for clinical review, not independently diagnosed.
- **Does an AI flag mean cancer?** No. A flag identifies a finding for closer review; the care team makes the diagnosis using the complete clinical picture.
- **Who should discuss pancreatic cancer screening?** People with a family history or known hereditary syndrome should discuss risk assessment with a specialist. Asymptomatic average-risk adults should not undergo routine pancreatic cancer screening.
- **Does the evidence guarantee an individual outcome?** No. Sensitivity, specificity and AUC describe defined study populations and test designs, not an individual diagnosis or treatment outcome.
- **Where is PANCREASaver available?** It is used in medical institutions and health checkup centers in Taiwan; public information is summarized on the For Health Systems page.

## Sources
- [USPSTF pancreatic cancer screening recommendation](https://www.uspreventiveservicestaskforce.org/uspstf/recommendation/pancreatic-cancer-screening)
- [Lancet Digital Health study, PMID 33328124](https://pubmed.ncbi.nlm.nih.gov/33328124/)
- [Radiology study, PMID 36098642](https://pubmed.ncbi.nlm.nih.gov/36098642/)

## Machine-readable discovery
- [Sitemap index](https://www.pancad.ai/sitemap_index.xml) and [canonical URL sitemap](https://www.pancad.ai/sitemap.xml) identify the public pages.
- [auth.md](https://www.pancad.ai/auth.md) documents the public, unauthenticated site-health endpoint only; it exposes no clinical, personal, or commercial data.