首創腎臟超音波預測慢性腎病之人工智能，台灣首篇發表於Nature Digital Medicine國際知名期刊
今周刊專訪「庶民AI大爆發」封面故事系列 (第1153期; 出刊日2019.01.28)
The Big Data Center of CMUH holds the high-quality medical data, including structured and unstructured data, and is capable to expend the value of these data by integratingit to the universal Taiwan’s National Health Insurance Database, the government-based registry data,and the environmental data.With these high dimensional and high-quality data, we can extract and generate real-world evidence for clinical practice and health policy-making.
Take the renal medicine at CMUH as examples. Wenow can use the serial repeated measurement ofeGFR to track the disease course and predict patients’ risk of developing adverse outcomes for each patient who received care at our hospital.
One example is the CKD Vigilant Information System. We found the first year eGFR variability is very useful in predicting the disease progression, CVD, and mortality for patients receiving CKD care in our hospital. Wedevelop a real-time automated risk alarming system to help nephrologist’s everyday practice.
Another example is the Acute Kidney Injury Detection System (AKIDS). We link our on-sitedata with the on-cloud national datatoobtain the complete health information. Then we develop a detection algorithm to screen forAKI and make referral recommendation for every outpatient in our hospital.Our system can also help identify the window for clinicians to avoid nephrotoxic agents in patients with AKI.