Media

2025-07-21

Excellence in AI Healthcare Technology Showcase in Taiwan: Tri-Service General Hospital, Taipei Veterans General Hospital, Taichung Veterans General Hospital, and Kaohsiung Medical University — Half Focusing on Cardiology and Critical Care


工作與社交網絡幸福感中取得完美平衡

Acura AKI, an AI-powered medical software developed by Taichung Veterans General Hospital, receives TFDA SaMD certification

Dr. Chun-Te Huang from the Department of Digital Medicine and Artificial Intelligence, Taichung Veterans General Hospital, presented Acura AKI, an AI-driven medical software. Leveraging federated learning and real-time electronic medical record data, the system enables real-time prediction of acute kidney injury (AKI) risk in clinical settings, further promoting smart healthcare and international collaboration in software as a medical device (SaMD).

Dr. Huang, a nephrologist and intensivist with over ten years of experience in internal and surgical ICUs, pointed out that the incidence of AKI in ICU patients is as high as 30–60%. Once AKI occurs and nephrologists are consulted, patient outcomes are often poor. Early identification of high-risk patients and timely interventions could significantly improve prognoses.

The team developed the system using limited clinical parameters—including blood pressure, heart rate, oxygen saturation, blood test results, and medication data—to perform hourly inference and predict a patient’s 24-hour AKI risk. The software has been granted TFDA SaMD certification.

The predictive model was initially built on the hospital’s internal database (2015–2020) and later externally validated with over 30,000 patient records from three major medical centers (Mackay Memorial Hospital, National Cheng Kung University Hospital, and Kaohsiung Medical University Hospital). A prospective randomized clinical trial is currently underway.

Dr. Huang emphasized that Acura AKI adopts federated learning to ensure patient privacy and data security, while enhancing the model’s generalizability and accuracy. Looking ahead, the team aims to collaborate with healthcare institutions in the EU to advance AI-powered smart healthcare and expand into the global market.

Previous Back Next