專題討論8:人工智慧在台灣的醫療應用

S8-1
Build a Smart Data Platform for Medical Intelligent Application: CMUH Experience to Ignite Hyper-Intelligence (iHi) Ecosystem
郭錦輯主任
中國醫藥大學附設醫院大數據中心

  High quality database and data flow management is the fundamental of “patient-oriented learning health system” in clinical medicine. Through assiduous data deep cleaning and analysis of electronic medical records, we are able to establish a research pipeline that can effectively clean structured and unstructured medical data, connect hospital-based data to government-based open data and claims data, aiming to precisely describe the disease course of each patient who ever sought medical care at our hospital. Using these high-quality longitudinal data, physicians can track the course and predict the possibility of developing adverse outcomes. In the end of 2020, we have launched iHi (Ignite Hyper-Intelligence) Platform at CMUH that allows researchers to create their own research dataset and perform deep analysis remotely. This “Data LEGO” approach tackles the common data bottlenecks ranging from data understanding among computing scientists, statisticians and health care professionals to hypothesis generation. The iHi Platform fills the last piece of CMUH Smart Data Ecosystem.