Technical Name SentinEID: An End-to-End, Genomics-to-Bedside AI Platform for Streamlined Pandemic Biosurveillance and Precision Clinical Decision Support
Project Operator National Taiwan University College of Medicine
Project Host 李建璋
Summary
SentinEID is an end-to-end "genomics-to-bedside" AI platform for pandemic defense. It integrates an optimized local IDseq pipeline with Taiwanese epidemiological databases to automatically streamline raw mNGS signals into clinical data. Powered by an edge-deployed MedGemma LLM, Knowledge Graphs, and RAG, it tethers to authoritative guidelines to deliver immediate, evidence-based antibiotic recommendations. Built on SMART standards, it is ready to launch on Taiwan's healthcare marketplace.
Scientific Breakthrough
This technology bridges the gap between high-dimensional genomic data and frontline care. By integrating an on-premise IDseq pipeline with a localized MedGemma LLM, it leverages Knowledge Graphs and RAG to translate complex mNGS signals into hallucination-free, evidence-based clinical insights in seconds. Standardized via SMART on FHIR, it establishes a new paradigm for rapid pandemic defense.
Industrial Applicability
This technology targets the smart epidemic prevention and precision diagnostics markets. It facilitates rapid tracking and source tracing of emerging viral variants while providing AI-generated AMR alerts and clinical summaries. Ideal for hospital diagnosis, disease surveillance centers, and genomic testing services, the platform’s multi-modal integration and dual-language support significantly enhance diagnostic efficiency and global collaboration potential.
  • Contact
  • PO CHUN, LIAO