Transforming Healthcare with AI: TMU Healthcare System Leads the Way

Addressing Challenges with Innovative Solutions

TMU-H faces challenges such as an imbalanced nurse-to-patient ratio and heavy workloads for medical staff, managing more than 3,000 beds across four affiliated hospitals, two centers, and one institution. To address these issues, TMU launched the Healthcare Information System (HIS) 3.0 Upgrade Plan, introducing generative AI and rethinking workflows from a system-wide perspective.

With strong support from TMU Chairman Ray-Jade Chen and President Mai-Szu Wu, TMU began a cross-hospital information system integration project in 2023 and completed platform standardization across its three affiliated hospitals in March 2024. Built on TMU’s “One Campus” strategy—“one platform, shared resources, and unified data”—the project established a forward-looking information architecture that strengthens the foundation for smart healthcare development.

President Mai-Szu Wu emphasized that digital transformation is not simply about tools, stating:
“Future healthcare will revolve around digitization, precision, and regeneration. Digital tools are essential, but transformation remains the true goal. Only with a strategic approach can we redesign medical processes and elevate clinical quality.”

TMU also independently developed the TMU AI Integrated Platform X (TAIP-X), which connects the databases of all affiliated hospitals and creates a unified data structure tailored to TMU’s clinical environment. This platform enables real-time cross-hospital data sharing and moves TMU-H closer to a fully integrated smart healthcare ecosystem.

Enhancing Medical Care Efficiency with AI-Driven Handover Processes

In July 2024, TMU-H implemented Generative AI (GenAI) in frontline nursing workflows, incorporating Retrieval-Augmented Generation (RAG) to enhance accuracy and reliability.
Traditional nursing handover documentation—typically requiring 6 to 11 minutes per patient—can now be completed in about 10 seconds, as the GenAI–RAG system automatically synthesizes the latest structured and unstructured clinical data.

A nurse responsible for seven patients could previously spend nearly an hour on documentation. With the AI-generated handover report, this task is now completed almost instantly, freeing valuable time for direct patient care.

TMU-H has since expanded GenAI–RAG applications to nursing care plans, clinical summaries, and automated Q&A tools. These enhancements show how AI has become an active part of everyday clinical practice.

Realizing AI-Powered Collaboration

In collaboration with Google Cloud, TMU-H adopted MedLM, a healthcare foundation model, and integrated the Gemini AI chatbot to build an internal generative AI system. Its intuitive design allows nurses to adapt quickly and provide feedback for continuous optimization.

Beyond documentation, AI applications have been extended to:
• medical imaging interpretation
• medication safety oversight
• physiological data analysis
• Smart IV Pump remote infusion monitoring

These technologies not only strengthen clinical decision support but also elevate patient safety across TMU-H’s hospitals. Frontline staff have expressed that AI tools now feel like practical clinical partners rather than experimental technology.

Laying the Groundwork for Precision Medicine

TMU-H is advancing toward data-driven precision medicine by planning to integrate clinical, genetic, and environmental data into AI-powered predictive models for personalized treatment plans. Its partnership with Google Cloud supports ongoing innovation and accelerates TMU-H’s transition toward precision healthcare.

Looking ahead, TMU-H is actively pursuing HIMSS EMRAM Stage 7 and the Digital Health Indicator (DHI) assessment. TMU also plans to expand GPU infrastructure to support faster medical imaging diagnostics and the deployment of more GenAI applications.

Committed to leveraging AI technology, TMU-H is pioneering advancements in Taiwan’s healthcare system, delivering a better healthcare experience and setting a new benchmark for medical innovation.