Development of a Clinical Decision Support System for Diabetes Patients

• Main Authors: Chong-Yi Chen, 陳崇毅
• Other Authors: Ting-Ying Chien
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/x3777g

碩士 === 元智大學 === 資訊工程學系 === 107 === In Taiwan, diabetes has become one of the most common chronic diseases. In addition to causing serious harm to people's health, diabetes has also caused problems such as wasting many medical resources. Therefore, how to save the medical expenses while further improving the health of diabetic patients is the main problem to be solved in this thesis. The Clinical Decision Support System (CDSS) proposed by the thesis provides the Glycated hemoglobin (HbA1c) in diabetic patients predicted by our model to the physician to assist them to prescribe. The network architecture used is a deep neural network architecture based on Bidirectional Long Short-Term Memory (Bi-LSTM) model. The error between the HbA1c predicted by our architecture and the actual HbA1c measurement of the patient is generally smaller than that predicted by the Support Vector Regression (SVR). The future HbA1c of the patient can be predicted more accurately as a basis for assisting the physician to prescribe.