论文部分内容阅读
本文采用无创性Bayes准则多元逐步判别分析方法,对73例同期接受改良Bruce方案平板运动试验及选择性冠状动脉造影对照的住院病人,进行回顾性资料处理(SAS/PC 6.03版微机软件系统),并将所建立的拟合模型对29例患者进行前瞻性研究.结果显示该多变量分析方法在建立良好的预测模型和判定指标的筛选方面具有独特价值,提示多变量分析方法可显著提高运动试验诊断冠心病的敏感性和特异性.
In this paper, noninvasive Bayes criteria multivariate stepwise discriminant analysis was used to retrospectively process data of 73 inpatients undergoing modified treadmill exercise test and selective coronary angiography in the same period (SAS / PC 6.03 software system) The prospective study of 29 patients with the fitted model showed that this multivariate analysis was of particular value in establishing a good prediction model and in screening screening criteria, suggesting that multivariate analysis could significantly improve exercise testing Diagnosis of coronary heart disease sensitivity and specificity.