Analysis of Serum Tumor Marker CA19-9 in the Diagnosis and Prognosis of Pancreatic Cancer-A Business Intelligence Approach

• Main Authors: Yu-XiangXue, 薛宇翔
• Other Authors: Sheng-Tun Li
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/3y5fp4

碩士 === 國立成功大學 === 高階管理碩士在職專班(EMBA) === 105 === (1) Scope and Objectives: Pancreatic cancer is a highly malignant cancer, and it is usually symptomless in the early stage, and only becomes clinically evident until the end phase of the disease course. The aim of this research is to construct a set of online analysis processing system based on the concept of business intelligence, we can analyze the diagnosis and prognosis of pancreatic cancer from different aspects, analyze the influence of different stages of pancreatic cancer, and explore the relationship between CA19-9 and survival time. (2) Methods: 812 CA19-9 abnormal results and diagnosis, were collected from a regional teaching hospital in the southern Taiwan from 2010 to 2012. The relationship between the variables was analyzed by chi-square test and independent T-test. Linear regression and support vector regression were used to analyze the prognosis of the patients. (3) Results: The risk of pancreatic cancer is increased in those patients with abnormal high CA19-9 value. Analysis of CA19-9 results and pancreatic cancer staging and location of the association, CA19-9 result can be found in the value of the pancreatic body tumor as the highest, the lowest for pancreatic head cancer. If the stage of pancreatic cancer, then the highest for the third period, the first phase of the lowest. (4) Conclusions: Based on the above analysis, we can found that CA19-9 values are not specific for the diagnosis of pancreatic cancer. If CA19-9 value is abnormal high, it can combine CT, PET-CT or MRI and other imaging tools to help diagnosis, and improve the diagnostic accuracy, In order to achieve the purpose of early diagnosis and early treatment.