進階搜尋


 
系統識別號 U0026-0812200910355160
論文名稱(中文) ROC曲線下面積相似性和非劣性之研究
論文名稱(英文) On the Equivalence and Non-inferiority for the Areas of Two ROC Curves
校院名稱 成功大學
系所名稱(中) 統計學系碩博士班
系所名稱(英) Department of Statistics
學年度 91
學期 2
出版年 92
研究生(中文) 陳伯辰
研究生(英文) Po-Chen Chen
電子信箱 pcchen@email.stat.ncku.edu.tw
學號 r2690102
學位類別 碩士
語文別 英文
論文頁數 95頁
口試委員 指導教授-馬瀰嘉
口試委員-戴政
指導教授-劉仁沛
口試委員-王新台
中文關鍵字 模擬  非劣性  相似性  ROC曲線  診斷檢定 
英文關鍵字 equivalence  non-inferiority  ROC curve index  ROC curve  simulation  diagnostic of test 
學科別分類
中文摘要 對於ROC曲線指標,過去大部分的文獻都是在探討兩種或是兩種以上診斷方法在相等性上的檢定。然而,此種檢定方法對於我們所關心的問題並非全然適當,因此就有些學者提出了相似性及非劣性的檢定方法以解決所要面對的問題。例如:在醫學上一種新的非侵入性診斷方法與一種已被認定為標準的侵入性診斷方法做比較,也許在診斷的效率上並不一定相等,但也許可以經過此種相似性及非劣性的檢定,得到新方法與標準的診斷方法是相似的,因此在成本、使用方便性及安全性的考量之下可以以新方法來代替標準的方法。對於ROC曲線指標而言,相似性或非劣性檢定所做的研究很少,本篇論文是利用DeLong, DeLong, Clarke-Pearson (1988) 無母數的方法,以及Dorfman and Alf (1968) 和Obuchowski and McClish (1997) 所提出的最大概似法來估計ROC曲線下面積,將ROC曲線下面積視為參數提出相似性及非劣性的檢定方法,並以模擬結果比較這三種方法的優劣。
英文摘要 In the past, for ROC curve index, most issue has focused on the question of whether the accuracy of two diagnostic tests differs. It may not be an appropriate question of interest in all situations, however. Hence, equivalence/non-inferiority test has been proposed to solve the questions of interest, e.g. in comparing diagnostic efficacy of an non-invasive alternative diagnostic (test) procedure to an invasive (reference) method. If the non-invasive alternative procedure is equivalence to the invasive method, we may use the non-invasive alternative diagnostic procedure because of its easy administration, its better safety profile or its reduced cost. But for ROC curve index, the literature on equivalence/non-inferiority test is scarce. In this paper, we compare the equivalence/ non-inferiority tests based on three methods for estimation of the area under ROC curve by DeLong, DeLong, and Clarke-Pearson (1988), Dorfman and Alf (1968), and Obuchowski and McClish (1997). A simulation study was conducted to empirically investigate the size and power of three methods.
論文目次 Chapter 1 Introduction………………………………………………..1
1.1 The ROC Curve………………………………………………………….4
1.2 The Area under the ROC Curve (AUROC)……………………………6
1.3 Tests for Equivalence or Non-inferiority……………………………......7

Chapter 2 Current Statistical Methods for Equivalence/Non-inferiority
Test of Two AUROCs.........................................................10
2.1 Interval Hypotheses for Two AUROCs………………………………10
2.2 Asymptotic Tests………………………………………………………11
2.3 Estimation of AUROC and Variance…………………………………12
2.3.1 Single ROC curve………………………………………………………………12
2.3.2 k correlated ROC curves……………………………………………………...….14

Chapter 3 Proposed Methods………………………………………..16
3.1 Analysis of AUROC under Binormal Model…………………………...16
3.1.1 AUROC under Binormal Model………………………………………………….16
3.1.2 Maximum Likelihood Estimation for Estimations of Parameters a and b……….17
3.2 Estimation of the Variance and Covariance of AUROC under
Binormal Model………………………………………………………...21
3.2.1 Variance and Covariance of AUROC under Binormal Model…………………...21
3.2.2 Variance and Covariance for Estimates of a and b................................................22
3.2.2.1 Single AUROC………..……………………………………………………………..22
3.2.2.2 Two Correlated AUROCs…………………………………………………………...23
3.3 Equivalence and Non-inferiority Tests Based on the parameters of Binormal Model and Between Two AUROCs under the Binormal
Model………………………………………………………..…………24

Chapter 4 Simulation study………………………………………….27
4.1 Simulation Process……………………………………………………..27
4.1.1 Data Generation………………………...……………………………………..…29
4.1.2 Selection of Cutoff Points and Initial Points……………………….…………….30
4.2 Simulation Results……………………………………………………..31
Chapter 5 Conclusion………………………………………………...36

Reference……………………………………………………………..38
Appendix A…………………………………………………………...41
Appendix B…………………………………………………………...54
Appendix C…………………………………………………………...56
參考文獻 [1] Bamber, D. (1975). The area above the ordinal dominance graph and the area below the receiver operation graph, J. Math. Psych. 12: 387-415.

[2] Blackwelder, W. C. (1982). Proving the null hypothesis in clinical trials, Controlled Clinical Trials 3: 345-353.

[3] Dorfman, D. D. and Alf, Jr., E. (1968). Maximum-likelihood estimation of parameters of signal-detection theory—a direct solution, Psychometrika 33: 117-124.

[4] Dunnett, C. W. and Gent, M. (1977). Significance testing to establish equivalence between treatments with special reference to data in the form of 2 2 tables, Biometrics 33: 593-602

[5] DeLong, E., DeLong, D., and Clarke-Pearson, D. (1988). Comparing the areas under two or more correlated receiver operation characteristic cures: A nonparameteric approach, Biometrics 44: 837-845.

[6] Farrington, C. P. and Manning, G. (1990). Test statistics and sample size formulae for comparative binomial trials with null hypothesis of non-zero risk difference of non-inferiority relative risk, Statistics in Medicine 9: 1447-1454.

[7] Feinstein, A. R. (2002). Principles of Medicial Statistics, Chapman and Hall/CPC, Boca Raton. FL.

[8] Hoeffding, W. (1948). A class of statistics with asymptotically normal distribution. Annals of Mathematical Statistics 19: 293-325.

[9] Hanley, J. A. and McNeil, B. J. (1982). The meaning and use of the area under a receiver operation characteristic (ROC) curve, Radiology 143: 29-36

[10] Lu, Ying and Bean, J. A. (1995). On the sample size for one-sided equivalence of sensitivities based upon McNemar’s test, Statistics in Medicine 14: 1831-1839.

[11] Luchenbruch, P. A. and Lynch, C. J. (1998). Assessing screening tests: extensions of McNemar’s test, Statistic in Medicine 17: 2207-2217.

[12] Liu, J. P., Hsueh, H. M., Hsieh, E., and Chen, J. J. (2002). Tests for equivalence or non-inferiority for paired binary data, Statistics in Medicine 21: 231-245.

[13] McNeil, B. J. and Adelstein, S. I. (1976). Determining the value of diagnostic and screening tests, J. Nucl. Med. 17: 439-448.

[14] Metz, C. E. (1978). Basic principles of ROC analysis. Seminars in Nuclear Medicine Ⅷ, 283-298.

[15] Metz, C. E. (1986). ROC methodology in radiologic imaging, Invest. Radiol. 21: 720-733.

[16] Metz, C. E. (1989). Some practical issues of experimental design and data analysis in radilolgic ROC studies, Invest. Radiol. 24: 234-245.

[17] Nam, J-m. (1997). Establishing equivalence of two treatments and sample size requirements in matched-pairs design, Biometrics 17: 1422-1430.

[18] Obuchowski, N. and McClish, D. (1997). Sample size determination for diagnostic accuracy studies involving binormal ROC curve indices, Statistics in Medicine 16: 1529-1542.

[19] Obuchowski, N. (1997). Testing for equivalence of diagnostic test, Am. J. Radiol. 168: 13-17

[20] Sen, P. K. (1960). On some convergence properties of U-Statistics. Calcutta Statistical Association Bulletin 10: 1-18.

[21] Sox, Jr., H. C., Blatt, M. A., Higgins, M. C., and Marton, K. I. (1989). Medical decision making, Batterworths-Heinemann, Boston.

[22] Tu, D. (1997) A comparative study of some statistical procedures in establishing therapeutic equivalence of nonsystemic drugs with binary endpoints, Drug Information Journal 31: 1291-1300.


[23] Yerushalmy, J. (1947). Statistical problems in assessing methods on medical diagnosis, with special reference to X-ray technique, Pub. Health Research 62: 1432-1449.

[24] Zhou, X. H., Obuchowski, N. A., and McClish D. K. (2002). Statistical Methods in Diagnostic Medicine, John Wiley and Sons, New York.
論文全文使用權限
  • 同意授權校內瀏覽/列印電子全文服務,於2003-06-16起公開。
  • 同意授權校外瀏覽/列印電子全文服務,於2003-06-16起公開。


  • 如您有疑問,請聯絡圖書館
    聯絡電話:(06)2757575#65773
    聯絡E-mail:etds@email.ncku.edu.tw