文档介绍:1 《中国循证医学杂志》 2004 年4卷2期起止页码: 132-135,103 ROC 曲线软件在处理不同类型资料中的应用探讨牛华肖质刘莉孙鹥王建琼黄永红台虹赵崇吉云南省临床检验中心(650032 昆明) 摘要:目的探讨在进行 ROC 曲线分析过程中对不同类型资料处理时应注意的问题及其解决方法。方法用云南省临床检验中心开发的中文 ROC 曲线分析软件,根据各心肌损伤标志物心肌肌钙蛋白( cTNT )、肌红蛋白( MYO )和肌酸激酶同工酶质量( CK-MB mass )测定结果资料分布特点及类型选择在全程数据范围内(即两组资料从最低值至最高值范围)或选择在重叠范围内进行 ROC 曲线分析处理。结果各心肌损伤标志物在不同时间内原始测定结果的资料分布极差非常大,经对数转换后资料分布极差明显减小可直接进行全程数据范围的 ROC 曲线分析,否则分析的结果无效或很不理想;若选择用两组资料重叠范围内的处理方法,其分析结果与对数转换后进行的全程数据范围处理结果一致。结论在进行 ROC 曲线分析时,应注意资料数据的分布规律,选择不同的统计方式。关键词: ROC 曲线中文软件心肌标志物对数转换 The application of ROC program with different types data Niu Hua, Xiao Zhi, Liu Li, Sun Yi, Wang Jianqiong, Huang Yonghong, Tai Hong, Zhao Chongji. Yunnan Provincial Clinical Laboratory Center, Kunming, 650032, China. 【 Abstract 】 Objective To explore the solutions of problems with the ROC analysis for different types data. Methods Two kinds of ROC analyses of three cardiac infarction markers, cTNT, MYO and CK-MB mass, were performed with the ROC program developed by Yunnan Provincial Clinical Laboratory Center. Results The distribution of prime data had a large range, which produce da bad analysis result. After logarithmic transformation, the prime data that had smaller range now can be analyzed with full-span method. The results were similar to the ROC analyzed with the overlap ped data. Conclusion We should choose different statistical method depend on the distribution of data when we pe