癌症，近年屢居全球人口死亡原因之冠，不論是什麼樣的癌症，往往造成許多家庭的破碎，到目前為止，還沒有找到一種可以完全治癒癌症的藥物。
本論文主要在卵巢癌的範疇上做討論，去探討在還未開刀前，如何可以準確的判斷病人的嚴重性，給予開刀的醫生足夠的資訊，以達到提升開刀成功率的目的，使用區別分析法(Discriminant Analysis)作為分類的依據，利用此方法來區分卵巢癌是惡性腫瘤(Invasive)或是中度惡性腫瘤(Borderline)，希望能夠把判斷的準確度提高。
以及在開完刀後，利用Cox 比例風險模型為基準，利用惡性腫瘤病人在手術之後的資訊，去分析其在不同期數與使用不同的化療藥物之下，對於存活風險的關係。

Cancer is one of the leading causes of death worldwide in recent years. Ovarian cancer is the second leading cause to death in women’s gynecologic cancers. Treatment of cancers is time- and resource-consuming. So far, there doesn’t exist any magical cure for cancers.
This study discusses main risk factors of ovarian tumors. One objective of this study was to build a model to differentiate borderline and invasive ovarian tumors. We hope to provide more information to doctors before the operation. We performed a retrospective study of patients with borderline or invasive ovarian tumors between 1997 and 2012 in Kaohsiung Veterans General Hospital. Clinical and pathological data were collected. Discriminant analysis was used to build a model in order to increase the accuracy of differentiation.
Another objective of this study was to analyze the risk factors for survival in patients with invasive ovarian tumors. We use the Cox proportional hazard model to analyze the risk of different stages and drugs.

誌謝i
摘要ii
Abstract iii
1 研究動機與目的1
2 資料描述3
2.1 資料收集. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.2 資料處理. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.3 變數介紹. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3 研究方法10
3.1 二次區別分析. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.2 彈性區別分析. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.3 Cox 比例風險模型. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.4 判斷準則. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.5 交叉驗證. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4 研究結果13
4.1 中度惡性腫瘤與惡性腫瘤的區別. . . . . . . . . . . . . . . . . . . . . . . . . 13
4.1.1 腫瘤大小的迷失. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4.1.2 選值判別法. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.1.3 區別分析. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.1.4 判斷準則. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
4.2 存活分析. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.2.1 期數風險比較. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.2.2 用藥風險比較. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.2.3 新藥與舊藥的藥效差異. . . . . . . . . . . . . . . . . . . . . . . . . . 21
5 討論與結語26
參考文獻27
附錄29

[1] Chang, C., Chow, S. N., and Hu, Y. (1995). Age of menopause of Chinese women in Taiwan. International journal of gynaecology and obstetrics: the official organ of the International Federation of Gynaecology and Obstetrics 49: 191-192.
[2] Curtin, J. (1994). Management of the Adnexal Mass. Gynecologic oncology 55: S42 - S46.
[3] Exacoustos, C., Romanini, M. E., Rinaldo, D., Amoroso, C., Szabolcs, B., Zupi, E., and Arduini, D. (2005). Preoperative sonographic features of borderline ovarian tumors. Ultrasound Obstet Gynecol 25: 50–59.
[4] Fruscella, E., Testa, A. C., Ferrandina, G., De Smet, F., Van Holsbeke, C., Scambia, G., Zannoni, G. F., Ludovisi, M., Achten, R., Amant, F., Vergote, I., and Timmerman, D. (2005). Ultrasound features of different histopathological subtypes of borderline ovarian tumors. Ultrasound in obstetrics & gynecology : the official journal of the International Society of Ultrasound in Obstetrics and Gynecology 26: 644-650.
[5] Geisser, S. Predictive inference : an introduction. (Chapman & Hall, 1993).
[6] Horvath, L. E., Werner, T., Boucher, K., and Jones, K. (2013). The relationship
between tumor size and stage in early versus advanced ovarian cancer. Med Hypotheses.
[7] Hastie, T., Tibshirani, R., and Buja, A. (1994). Flexible Discriminant Analysis by Optimal Scoring. Journal of the American Statistical Association 428: 1255-1270.
[8] Kleinbaum, David G. (2005). Survival analysis : a self-learning text, 2nd ed. Springer, New York.
[9] Morotti, M., Menada, M. V., Gillott, D. J., Venturini, P. L., and Ferrero, S. (2012). The preoperative diagnosis of borderline ovarian tumors: a review of current literature. Archives of gynecology and obstetrics 285: 1103-1112.
[10] Moore, R. G., Miller, M. C., Disilvestro, P., Landrum, L. M., Gajewski, W., Ball, J. J., and Skates, S. J. (2011). Evaluation of the diagnostic accuracy of the risk of ovarian malignancy algorithm in women with a pelvic mass. Obstetrics and gynecology 118: 280-288.
[11] Nossov, V., Amneus, M., Su, F., Lang, J., Janco, J. M., Reddy, S. T., and Farias- Eisner, R. (2008). The early detection of ovarian cancer: from traditional methods to proteomics. Can we really do better than serum CA-125? American journal of obstetrics and gynecology 199: 215-223.
[12] R: A language and environment of statistical computing (R Foundation for Statistical Computing, Vienna, Austria, 2010).
[13] Rencher, Alvin C. (2002). Methods of Multivariate Analysis, 2nd ed. Wiley, New York.
[14] Sobiczewski, P., Rzepka, J., Gujski, M., Michalski, W., Da´nska-Bidzi´nska, A., Kupryja´nczyk, J., and Bidzi´nski, M. (2012). Evaluation of selected ultrasonographic parameters and marker levels in the preoperative differentiation of borderline ovarian tumors and ovarian cancers. Archives of gynecology and obstetrics 286: 1513-1519.
[15] Sassone, A. M., Timor-Tritsch, I. E., Artner, A., Westhoff, C., and Warren, W. B. (1991). Transvaginal sonographic characterization of ovarian disease: evaluation of a new scoring system to predict ovarian malignancy. Obstetrics and gynecology 78: 70-76.
[16] Trope, C., Davidson, B., Paulsen, T., Abeler, V. M., and Kaern, J. (2009). Diagnosis and treatment of borderline ovarian neoplasms ”the state of the art”. European journal of gynaecological oncology 30: 471-482.
[17] Zanetta, G., Rota, S., Chiari, S., Bonazzi, C., Bratina, G., and Mangioni, C. (2001). Behavior of borderline tumors with particular interest to persistence, recurrence, and progression to invasive carcinoma: a prospective study. Journal of clinical oncology : official journal of the American Society of Clinical Oncology 19: 2658-2664.
[18] 吳維文(2012)。南台灣醫學中心關於卵巢腫瘤風險因子研究，國立中山大學應用數學系碩士論文。