摘要
目的分析12项肿瘤标志物在前列腺癌中的表达,进而运用Bayes法(贝叶期法)建立多肿瘤标志物前列腺癌判别函数,探讨Bayes法在前列腺癌诊断中的作用和临床意义。方法用蛋白芯片法检测2177例恶性肿瘤患者(其中21例前列腺癌)和2111例正常及良性病变者的12项常见肿瘤标志物,应用Bayes法建立肿瘤三级判别诊断函数。结果(1)一级判别函数对前列腺癌诊断准确率为83.97%,灵敏度为71.28%,特异度为82.11%。二级判别函数对前列腺癌诊断诊断的准确率为96.87%,灵敏度为93.33%,特异度为100%。(2)三级诊断判别函数对前列腺癌诊断的准确率为81.82%,部分前列腺癌被误诊为食管癌。(3)成功建立了新的多肿瘤标志物流程诊断软件。结论基于Bayes法建立诊断判别函数能显著提高前列腺癌诊断,具有较高的临床应用价值。
Objective To analyze the expression of C-12 multiple tumor marker in prostate cancer,futhermore,to establish the diagnostic function by Bayesian methods and to evaluate the function of bayesian methods and clinical values in diagnosis of prostate cancer. Methods 12 tumor markers were detected by protein chip technology in 2 177 cases of malignant tumors （prostate cancer 21 cases ） and 2 111 cases in control group. To retrospectively study their clinical information, furthermore to establish third grades diagnostic function by Bayesian method. Results （1） The accuracy rate, sensitivity and specificity were 83. 97% ,71. 28% and 82.11% by the first grade diagnostic function and 96.87 %, 93.33 % and 100% by the second diagnostic function, respectively. （2） The accuracy rate was 81.82 % by the third grade diagnostic function. Part of prostate cancer was diagnosed esophageal carcinoma. （3）The multiple tumor marker analysis system software was developed successfully. Conclusion The diagnostic function shows great potential for the early detection of prostate cancer. The function is helpful for clinical use and points out new diagnostic trend for prostate cancer.
出处
《重庆医学》
CAS
CSCD
北大核心
2010年第5期527-529,532,共4页
Chongqing medicine
关键词
多肿瘤标志物
前列腺癌
贝叶斯法
软件
诊断函数
multiple tumor marker
prostate cancer
Bayesian methods
software
diagnostic function
