在現今的醫療技術大幅進步之下，子宮內膜異位症的病患很多可經由手術來治癒，讓患者可以懷孕生育。在醫療統計上，許多臨床醫師和學者證實了“超音波導引抽取注入95%酒精7分鐘以上至不抽出的治療方式有較佳的療效”，但是卻很少學者探討“不同狀況患者對酒精治療時間長短與病患疼痛改善率之間的關係。基於不同狀況病患的療效會有所不同，術後疼痛的改善與否也會因不同的治療方式而有所不同，所以本論文以資料探勘方式探究何種治療方式對病患疼痛有較佳的改善效果。因為病患資料欄位較多，傳統僅以統計方法t檢定或變異數分析等方式，不易找到合適的切點去分割成適合資料群組來分析，因此本研究採用決策樹演算法並結合統計方式t檢定，分析治療方式對疼痛有改善與沒改善的人數，並比較各治療方式對患者的疼痛改善情況，藉以了解何種治療方式具有較佳的疼痛改善療效。除了當囊腫數量大於4時，在其他狀況下，都是酒精滯留7分鐘以上的治療方式具有顯著的疼痛改善療效。事實上，除了樣本數不足的問題外，經與專業醫師詢問後，發現在囊腫數量較多的病況下，子宮內膜異位症的患者不容易治癒，因此不論使用那一種治療方式，對患者疼痛的改善效果也是有限的。

Nowadays medical technology has improved rapidly; many patient of endometriosis can be treated with surgery, which enables the patient to be pregnable. Medical statistics data indicates many clinicians and researchers have confirmed that the therapy of ultrasound-guided aspiration with 95% ethanol for up to 7 minutes without extraction has the best result for ovarian endometrioma, however, only few of them have explored the relationship on the length of ethanol treatment time for patient with different conditions and the pain relief improvement rate of the patient. The effectiveness of the treatment is different in accordance to different patient condition; and the pain relief improvement rate is different in accordance to different treatments. This study implements data mining approach to explore the most effective treatment methods for patient pain relief. Due to the fact that there are too many data fields on patients, by applying the traditional statistical methods, like t-test or analysis of variance (ANOVA) methods would be difficult to find a suitable cut-off point to be divided into suitable groups for data analysis. Therefore, this study uses the decision tree algorithms to analyze the number of patient for pain-scale have improved or not in different treatment, and combined with statistical methods t-test, to compare the treatment of patients with pain improvement rate in order to understand what kind of treatment has better in pain-scale relief effect. Furthermore, compare the patient pain improvement condition on different treatments and understand what the most effective treatments in pain relief are. Ethanol for up to 7 minutes without extraction has the best result in pain relief except when the number of cyst exceed 4. In fact, after consulting with physicians the finding is that apart from lack of sample issue, endometriosis is difficult to be treated when large number of cyst is present, thus, regardless of which treatment is applied, the effectiveness is limited.

目  錄
第1章	緒論	1
1.1	研究背景與目的	1
1.2	論文架構	4
第2章	相關知識背景	5
2.1	子宮內膜異位症簡介	5
2.1.1	子宮內膜異位症的症狀	7
2.1.2	子宮內膜異位症的檢查方式	9
2.1.3	子宮內膜異位症的治療方式	10
2.2	資料探勘	12
2.3	決策樹演算法	15
2.3.1	決策樹演算法流程	16
2.3.2	CART分類與回歸樹	19
第3章	研究方法	21
3.1	流程架構及研究方法	22
3.2	資料預處理與說明	25
第4章	實驗結果與分析	28
4.1	實驗環境與目標	28
4.2	基本分析	29
4.3	酒精滯留時間與疼痛改善的情形	33
第5章	結論與未來研究方向	40
參考文獻	41
附錄1-中文論文摘要	45
附錄2-英文論文	67
 
圖目錄
圖 2.1  子宮內膜異位症的發生部位[2]	6
圖 2.2  子宮內膜異位症增生位置與女性子宮構造圖	8
圖 2.3  知識發掘流程圖（KDD process）	12
圖 2.4  預測客戶是否購買NB雷腦的決策樹	15
圖 2.5  建構決策樹的基本演算法	17
圖 3.1  研究方法流程圖	22
圖 4.1  分析病患疼痛改善情況的決策樹	29
圖 4.2  病患酒精滯留時間與疼痛改善情況的決策樹	35

 

表目錄
表 3.1  分析的欄位資料說明	25
表 3.2  標識用衍生屬性表	26
表 4.1  依圖4.1比較7分鐘以上和7分鐘以下對疼痛改善比率的分析	31
表 4.2  依圖4.2 分析同一病症下治療方式與疼痛改善的關係	38

[1] 稀欀愀渀 S, Arici A. Advances in treatment options of endometriosis. Gynecologic and obstetric investigation 2008;67:81-91.
[2] 劉琦殷. 子宮內膜異位症. 嘉基藥訊 2015:4-6.
[3] 張夏欣. 經痛、不孕、內膜異位－認識子宮內膜異位症. 中華民國內膜異位症婦女協會會刊 2007;14:11-3.
[4] 曾仁宇. 子宮內膜異位. 中華民國內膜異位症婦女協會會刊 2009;16:4-5.
[5] 編輯部. 子宮內膜異位症. 當代醫學 2012:142-.
[6] Aboulghar MA, Mansour RT, Serour GI, Rizk B. Ultrasonic transvaginal aspiration of endometriotic cysts: an optional line of treatment in selected cases of endometriosis. Human Reproduction 1991;6:1408-10.
[7] Noma J, Yoshida N. Efficacy of ethanol sclerotherapy for ovarian endometriomas. International journal of gynecology & obstetrics 2001;72:35-9.
[8] Hsieh C-L, Shiau C-S, Lo L-M, Chang M-Y. Effectiveness of ultrasound-guided aspiration and sclerotherapy with 95% ethanol for treatment of recurrent ovarian endometriomas. Fertility and sterility 2009;91:2709-13.
[9] 陳信孚. 子宮內膜異位症患者如何面對各項不孕症治療方式. 中華民國內膜異位症婦女協會會刊 2004;11:5-.
[10] 黃小玲, 何日生, 陳史清. 注射無水酒精法治療卵巢子宮內膜異位囊腫18例探討. 齊齊哈爾醫學院學報 2006.
[11] 曾雅潔, 馬澤麟. 注射無水酒精法治療卵巢子宮內膜異位囊腫的改進. 廣東醫學院學報 2000;18:299-300.
[12] Bolton S, Bon C. Pharmaceutical statistics: practical and clinical applications: CRC Press; 2009.
[13] Bolton S, Bon C. Linear Regression and Correlation Pharmaceutical statistics: practical and clinical applications: CRC Press; 2009.
[14] Wang H, Zhang P. A Quantitative Method for Pulse Strength Classification Based on Decision Tree. JOURNAL OF SOFTWARE 2009;4:323.
[15] Timofeev R. Classification and regression trees (CART) theory and applications: Humboldt University, Berlin; 2004.
[16] QUINLAN J. Induction of Decision Trees. Machine Learning 1986;1:81-106.
[17] 四季台安醫院【內視鏡手術中心】. 各式疾病簡介-子宮內膜異位症. 2014.
[18] 林淑玲, 夏萍, 李奇龍, 許樹珍. 子宮內膜異位症對婦女婚姻生活之影響. 新臺北護理期刊 2002;4:23-35.
[19] 張明揚醫師. 子宮內膜異位症的發生部位. 中華民國子宮內膜異位症婦女協會.
[20] 顏兆熊. 子宮內膜異位症. 當代醫學 2004:559-66.
[21] 何彥秉 台. 子宮內膜異位症是否需要開刀切除卵巢？.  台中榮總醫訊2012.
[22] 莊斯婷. 整合決策樹與統計分析超音波抽取注入酒精在子宮內膜異位症療效. 新北市: 淡江大學; 2010.
[23] 盧凡. 子宮內膜異位症的藥物治療. 中華民國內膜異位症婦女協會會刊 2001;8:9-10.
[24] 巢佳莉. 子宮內膜異位症診斷與藥物市場現況與研發趨勢. 馬偕醫護管理專科學校學報 2004:79-96.
[25] 羅光霞. 兩種途徑保守性手術聯合藥物治療子宮內膜異位癥的臨床分析. 數理醫藥學雜志 2010;23:49-50.
[26] Aboulghar MA, Mansour RT, Serour GI, Sattar M, Ramzy AM, Amin YM. Treatment of recurrent chocolate cysts by transvaginal aspiration and tetracycline sclerotherapy. Journal of assisted reproduction and genetics 1993;10:531-3.
[27] Koike T, Minakami H, Motoyama M, Ogawa S, Fujiwara H, Sato I. Reproductive performance after ultrasound-guided transvaginal ethanol sclerotherapy for ovarian endometriotic cysts. European Journal of Obstetrics & Gynecology and Reproductive Biology 2002;105:39-43.
[28] Suganuma N, Wakahara Y, Ishida D, Asano M, Kitagawa T, Katsumata Y, et al. Pretreatment for ovarian endometrial cyst before in vitro fertilization. Gynecologic and obstetric investigation 2002;54:36-42.
[29] Kafali H, Yurtseven S, Atmaca F, Ozardali I. Management of non-neoplastic ovarian cysts with sclerotherapy. International Journal of Gynecology & Obstetrics 2003;81:41-5.
[30] Fayyad UM, Piatetsky-Shapiro G, Smyth P, Uthurusamy R. Advances in knowledge discovery and data mining: AAAI press Menlo Park; 1996.
[31] Fayyad UM, Piatetsky-Shapiro G, Smyth P, Uthurusamy R. ADVANCES IN KNOWLEDGE DISCOVERY AND DATA MINING. 1998.
[32] FAYYAD U, PIATETSKY-SHAPIRO G, SMYTH P. From data mining to knowledge discovery in databases. The AI magazine 1996;17:37-54.
[33] Bellazzi R, Zupan B. Predictive data mining in clinical medicine: current issues and guidelines. International journal of medical informatics 2008;77:81-97.
[34] 丁一賢, 資訊管理, 陳牧言, 資訊管理. 資料探勘: 滄海; 2005.
[35] Quinlan J. Discovering rules form large collections of examples: A case study. Expert Systems in the Micro Electronic Age Edinburgh Press: Edingburgh 1979.
[36] Quilan J. Learning efficient classification procedures and their application to chess end games. Machine Learning: An Artificial Intelligence Approach 1983;1.
[37] Jiawei H, Kamber M. Data mining: concepts and techniques. San Francisco, CA, itd: Morgan Kaufmann 2001;5.
[38] Breiman L, Friedman J, Stone CJ, Olshen RA. Classification and regression trees: CRC press; 1984.
[39] Breiman L, Friedman J, Olshen R, Stone C. Classification and Regression Trees.  1984.
[40] Chang M-Y, Chiang R-D, Wu S-J, Chan C-H. Mining unexpected patterns using decision trees and interestingness measures: a case study of endometriosis. Soft Computing 2015:1-13.