Comparison of endometriosis treatment using decision tree and t-test

• Main Authors: Ping-Wei Huang, 黃炳維
• Other Authors: 陳俊豪
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/17345925500572552959

碩士 === 淡江大學 === 資訊工程學系碩士班 === 100 === According to statistical research studied by WHO, there’re approximately 20% of infertility female patients who had been affected by endometriosis. But these patients can be healed by medicated or surgery treatment due to improvement of medical technology. At present, “7 minutes or above” treat time became the main treatment if the injection of 95% ethanol was chosen to be the surgery treatment, meanwhile, doctors would use medicated treatment after surgery likely according to the patient’s status. However, the effect of both medicated and surgery treatment can be different due to the difference of patient’s status. Therefore, the goal of this paper is to find out which treatment would be better rest with patient’s certain status. If we only use statistical way (ANOVA, t-test, etc) to solve this problem, how to partition the medical data properly would be a serious circumstance. Therefore, CART decision tree was used to generate different patient’s status in this paper, then, compare population, cure rate and magnificent difference of t-test between different treatments, after that, the better medical treatment can be found. According to the experiment result in this paper, the better medical treatment can be found out by purposed method.