Hospitalized or emergency department patients
All ward patients
High-risk mortality patients
Rural area residents (88 service points)
Developed a proprietary AI-enabled electrocardiographic (ECG) mortality prediction model integrated with clinical trial validation. The AI alert system predicts patient mortality risk within 90 days, enabling physicians to perform early intervention and intensive care, demonstrating significant improvement in patient survival rates.
AI-ECG accuracy of 0.93-0.96, superior to international standards
World's first clinical trial validating AI-ECG mortality reduction
Overall mortality reduction of 17%, high-risk patient mortality reduction of 31%
Cardiac mortality reduction of 93%
Cost-effectiveness of 99.6%
Approved by Taiwan FDA as medical device
Published in Nature Medicine with impact factor 58.7
Clinical trial sample: 16,335 patients (AI alert intervention group 8,177, routine care group 8,158)
High mortality risk patients: 709 (8.9%)
Rural area ECG interpretation volume: over 29,000 cases
Hospital coverage: 415 attending physicians, 1,949 nurses, 780 administrative staff, 707 other healthcare professionals, 776 physicians
All-cause mortality reduction of 17% within 90 days
High-risk patient mortality reduction of 31%
Cardiac mortality reduction of 93%
Non-cardiac mortality reduction of 24%
Increased ICU admission rate, antiarrhythmic drug use, echocardiography rate with physician intervention
AI prediction accuracy: internal validation 0.93-0.96, external validation 0.88
Can detect 29 acute and chronic cardiovascular diseases, interpret 24 arrhythmia conditions
AI mortality risk prediction accuracy of 99%
Randomized clinical trial design ensures scientific rigor
Dual validation mechanism with internal data validation and external data verification
Approved by Taiwan FDA as medical device
Regular monitoring of AI model accuracy indicators by team
High physician acceptance of AI alert system as clinical reference
Early prognostic notification improves clinical decision-making confidence
Multiple medical centers actively adopting the system (Fuhsing Hospital, Far Eastern Memorial Hospital, etc.)
Published in Nature Medicine (impact factor 58.7, ranked first in medical research, out of 189 journals)
Published over 30 SCI international papers
Recognized and reported by American Heart Association
International multi-center validation collaboration: Cedars-Sinai Medical Center Los Angeles, Mount Sinai Center New York, Seoul National University Hospital, Oxford University Hospital
2022 Future Technology Award
2023 Taipei Biotech Award - Cross-domain Excellence Category Gold Award
2023 MediaTek Smart Hometown Competition Excellent Award
Cost-effectiveness analysis of 5000 cases: 99.6% probability of cost-effectiveness with AI-ECG intervention
Reduces in-hospital cardiovascular mortality events
Improves patient outcomes, enables early detection of potential cardiac diseases
Enhances physician clinical decision support
Advances Taiwan's smart healthcare development
Benchmark learning and multi-center verification domestically and internationally
Technology transfer collaboration with Quanta Services in 2023
