1. Agentic AI – An Intelligent Assistant to Reduce Clinical Workload
Healthcare professionals often spend significant time on administrative tasks and clinical documentation, limiting direct patient interaction. To address this challenge, FEMH developed its own “GenAI Smart Healthcare Platform” as a digital assistant for healthcare staff. The platform integrates Azure OpenAI with an in-house Retrieval-Augmented Generation (RAG) knowledge base, enabling rapid generation of admission notes, discharge summaries, and accurate retrieval of internal ISO/SOP guidelines. Currently, the platform supports approximately 10% of patient care workflows, significantly reducing documentation time and allowing healthcare professionals to focus more on patient care.
Ambient AI – Creating a Seamless Voice-Driven Documentation Environment
Traditional medical documentation relies heavily on keyboard input, which can reduce efficiency and disrupt physician–patient communication. In collaboration with the Industrial Technology Research Institute (ITRI) and Microsoft, FEMH introduced an “AI Smart Voice Medical Documentation System.” Using advanced speech recognition capable of understanding Mandarin, English, Taiwanese, and specialized medical terminology, the system converts physician–patient conversations into structured medical records in real time. This Ambient AI technology has shown particularly strong performance in asthma voiceprint recognition and has optimized 10–20% of inpatient documentation workflows, moving toward the goal of “speak-to-document” intelligent healthcare.
2. Administrative AI – Enhancing Operational Efficiency and Decision-Making
As hospital operations and administrative processes become increasingly complex, administrative staff must manage large volumes of documents and data. FEMH has comprehensively adopted AI tools to strengthen operational management, utilizing its internal GenAI platform for document drafting, meeting summarization, and legal contract review. In addition, Microsoft Copilot is used for data integration and KPI dashboard development. These AI applications currently support 20–50% of administrative workflows, reducing human error while enabling hospital leadership to make timely and data-driven management decisions.
3. Clinical Laboratory AI – Strengthening Diagnostic Accuracy and Early Warning Systems
Clinical laboratories process large numbers of specimens daily, making rapid and accurate identification of abnormal results essential for patient safety. At FEMH, AI-assisted applications in the Department of Clinical Pathology now exceed 80% coverage. Through systems such as the “AI Blood Analysis System” and mass spectrometry-based microbial identification, the hospital has significantly improved the accuracy of cell classification and bacterial identification. In addition, AI-powered abnormal value interception automatically flags suspicious reports, while pharmacogenomics applications help physicians avoid medication allergy risks, establishing a strong safety barrier within the laboratory diagnostic process.