Virtual Patient AISAC

The future of oral implantology is here—and it’s powered by intelligent decision support.

At EURODOC, we’re launching Virtual Patient AISAC, a groundbreaking clinical decision support system that validates AI-assisted diagnosis in implant dentistry for the first time. We’re proving that machine intelligence can match—and even enhance—expert clinical judgment.

Why This Matters

Traditional implant planning still depends heavily on each individual clinician’s risk classification, which means that two dentists can rate the same case very differently. This experience‑driven variability leads to inconsistent planning quality, unpredictable outcomes, and avoidable complications—especially in complex implant cases.

Virtual Patient AISAC Solution

✓ Automatically analyzing clinically relevant patient parameters

✓ Reducing diagnosis time from 20 minutes to just 5 minutes

✓ Standardizing risk assessment across all cases

✓ Supporting less experienced clinicians with expert-level diagnostics

✓ Maintaining human oversight—the AI never decides alone

The Science

Led by Dr. Michael Truppe, this prospective study will compare AISAC clinical decision support against blinded expert panels using rigorous statistical methods. We’re measuring diagnostic accuracy, system usability, patient satisfaction, and long-term implant outcomes including osseointegration and marginal bone loss.

• In 1999, Sailer and Pajarola introduced the first S / A / C scheme (Simple, Advanced, Complex) to grade surgical difficulty and give general dentists a practical risk compass for oral surgery.
• The ITI and national implantology societies then adopted and refined this into the ITI SAC classification, separating surgical and prosthetic difficulty, adding aesthetic and systemic risk factors, and eventually codifying it in Dawson & Chen’s 2009 SAC monograph as the global reference for implant complexity.
• MedlibreGPT and now Virtual Patient AISAC keep the ITI SAC logic but extend it into the digital era: more than 50 clinical, radiological, and systemic parameters are encoded in machine-readable decision trees, processed by ML models and agentic AI workflows, and combined with guideline-driven RAG to deliver standardized, explainable SAC and risk classifications—while clinicians stay in full control.

AI Supercomputer as Game Changer

Virtual Patient AISAC, although a High-Risk AI System under the EU AI Act, achieves full conformity with MDR and EU AI Act requirements because it runs on an on‑premises petaflop‑class AI supercomputer that keeps all data inside the clinic’s secure environment. By combining this infrastructure with validated, open‑source large language models, AISAC turns regulation into an innovation driver: it delivers transparent, explainable decision support where every inference is logged, fully traceable to its data sources, and continuously overseen by clinicians, while all processing remains GDPR/DSGVO‑compliant through local execution, strong encryption, and fine‑grained access control.

This regulated, on‑premises architecture is a game changer compared with generic cloud tools, which often end up as “shadow AI” operating outside formal hospital governance, data protection, and notified‑body oversight. Instead of asking clinicians to choose between performance and compliance, AISAC provides both: supercomputer‑level AI that meets the strictest European requirements for safety, transparency, and accountability in everyday clinical practice.

The vision? Democratizing expert-level implant diagnostics globally, one algorithm at a time.