基于粒子群和非参数检验的T波交替联合检测方法  

Method of T Waves Alternans Joint Detection Based on Particle Swarm and Nonparametric Test

作　　者：佘黎煌[1] 卢丽[1] 张石[1] 王明全[1] 

机构地区：[1]东北大学信息科学与工程学院,辽宁沈阳110819

出　　处：《东北大学学报（自然科学版）》2013年第3期326-329,共4页Journal of Northeastern University(Natural Science)

基　　金：中央高校基本科研业务费专项资金资助项目(N110404003);辽宁省自然科学基金资助项目(20102057)

摘　　要：为了提高TWA检测方法的准确性和鲁棒性,提出了基于粒子群与秩和检验的T波交替联合检测方法.通过基于高斯核的心电图模型,利用粒子群算法来最优提取和对齐T波,提高了T波提取的鲁棒性和可靠性,降低了由于畸变T波引起的TWA误检和漏检.同时,基于非参数方法来定量和定性计算TWA参数,由于无需估计TWA的概率模型,提高了对现实复杂TWA现象检测的鲁棒性.仿真实验表明,测量的幅度与参考的真实值的相关系数达到0.96.标准的TWA数据和实际的心电图数据测试表明,所提出的TWA检测方法具有较高的鲁棒性.In order to improve the accuracy and robustness of the TWA （T wave alternans） detection method, a method based on particle swarm and rank-sum test was developed to detect TWA. By using particle swarm optimal method to extract and align T waves, the accuracy and robustness of TWA detection could be improved and the false detection of TWA caused by distortion could be reduced. TWA parameters were quantitatively and qualitatively calculated on the basis of nonparametric method, and the robustness of the complex TWA phenomenon detection was improved because it was no need to estimate the probability model of the TWA. The simulation experiment showed that the correlation coefficient between the experimental results and the true value was 0.96. The test of the standard TWA data and the actual ECG （electrocardiograph） data indicated that the proposed TWA detection algorithm has a high robustness.

关 键 词：粒子群 秩和检验 T波交替 心电模型 心电图 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]