信号平均心电图频域分析方法的研究

A METHODOLOGICAL STUDY OF FREQUENCY DOMAIN ANALYSIS FROM THE SIGNAL-AVERAGED ELECTROCARDIOGRAM

对53例心肌梗塞不伴室速(Ⅰ组),27例心肌梗塞伴室速(Ⅱ组)和42例健康者(Ⅲ组)的信号平均心电图频域多变量分析,选择其最佳分析参数.结果表明,QRS始点后60ms分析段(A_(60))20～50Hz,50～80Hz的A谱面积比(AAR),20～50Hz的P谱面积比(PAR)三组间均有显著差异(P<0.001);Bayes,Fisher分析示:20～50Hz,50～80Hz的AAR为相对独立的判别参数,频域分析对判别心肌梗塞伴室速的正确率为85.00%,敏感度77.78%,特异性88.68%,阳性和阴性预测值分别为77.78%和88.68%.提示频域分析能准确分辨心肌梗塞伴室速病人.

The best parameters were studied by frequency domain analysis of signal-averaged electrocardiogram from 53 remote myocardial infarction (RMI) patients without ventricular tachycardia ( VT ) (Group I), 27 RMI patients with VT ( Group Ⅱ ) and 42 normal subjects (GroupⅢ).There were significant differences among the 3 groups for 20～50Hz, 50～80Hz A spectra area ratio (AAR) and 20～50Hz P spectra area ratio (PAR) using the interval of starting point of 60ms after the QRS onset(A60) ( P<0.001).Bayes and Fisher stepwise discriminant analysis showed that 20～50Hz, 50～80Hz AAR of A60 segment were a relatively indepentent paramenters. Frequency domain analysis provided a accuracy of 85%, a sensitivity of 77.78%, a specificity of 88.68%, a positive prognostic value of 77.78%, and a negative prognostic value of 88.68%. This suggested that patients of RMI with VT were correctly identified by frequency domain analysis, which is a better method for the analysis of ventricular late potentials.