AIDE leverages large language models and generative AI to integrate early vital signs and chief complaints of emergency patients. It enables early risk detection and real-time recommendations, improving care efficiency, reducing misdiagnosis and delays, and ensuring timely, accurate treatment while lowering unnecessary costs. Recognized with the 2025 AI Innovation Award, AIDE demonstrates strong potential for clinical decision support.

AIDE is Taiwan's first AI decision system for emergency care, achieving 0.87 AUROC in cross-hospital validation. It integrates CNTR for multilingual clinical text processing and handles multimodal data (vital signs, symptoms, wearables). AIDE supports multi-task triage predictions: critical illness, imaging needs, admission, and return risk. Key innovations include imbalanced data handling, local deployment, and full-cycle decision support, demonstrating strong clinical generalizability.

AIDE help hospitals reduce patient wait times, ensure care quality, address workforce shortages, enhance reputation and competitiveness, and strengthen system advantages. It adopts a revenue-sharing model with hospitals, requiring no high upfront investment and payment is based on actual returns, lowering adoption barriers and aligning stakeholder interests. Protected by strong intellectual property rights, AIDE offers immediate clinical utility and long-term commercial potential.