A medical decision support system using text mining to compare electronic medical records

• Main Authors: CHAO,CHING PING, 趙竟評
• Other Authors: Pei-Ju,Lee
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/mg95bs

碩士 === 國立中正大學 === 資訊管理系醫療資訊管理研究所 === 108 === With new technological advances, Electronic medical record has gained popularity. The electronic medical record brings many benefits such as convenient storage and the ability to copy and paste the medical record to speed up the enter electronic medical record. However, it also has some shortcomings, such as repeated postings caused the electronic medical records to be lengthy, making the medical staff less readable or pasting the wrong drug causing medical negligence. In the past studies, no research used Bigram model to find out the fixed compared medical records to find out the new medical information The purpose of this study is to use the Bigram comparison to highlight new information on electronic medical records. After pre-processing data to remove noise, Gold Standard of the electronic medical records defined by two expert physicians was compared to the highlighted lines of Bigram comparison system. The F-measure was used to find out a fixed number of medical records to compare the non-highlight electronic medical records to highlight new medical information on the web interface. The experiment results suggested using four medical records in the comparison. When the fixed medical records are compared with the medical records, the speed of comparison can be faster and more medical records can be marked. When comparing the medical record information, the user first determines the number of medical records and reduces the comparison time. Therefore, this system allows medical staff able to obtain faster decision making.