Nonlinear dynamic characteristics analysis of synchronous 12-lead ECG signals.

Abstract

The authors computed the correlation dimensions of the synchronous 12-lead ECG of 30 healthy persons and 30 coronary heart disease (CHD) patients. They found that the D/sub 2/ of the same person is different for different lead ECG signals. Therefore, it is not appropriate to describe the human heart, a nonlinear dynamic system with complex structures, by only a single fractal dimension. Synchronous 12-lead analysis of correlation dimension can present a more detailed description of the heart's dynamic characteristics than can a simple one-lead analysis. It is important to specify the particular lead position when one refers to the D/sub 2/'s of the ECG signal. The authors also find that the differences among D/sub 2/'s of 12-lead ECG signals are relatively small for healthy people, compared with the same lead for CHD patients. For CHD patients, the situation is complicated. The differences among D/sub 2/'s for 12-lead ECG signals of some patients are large. Correlation dimensions of chest-lead ECG signals are obviously smaller than those of limb-lead ECG signals. This suggests that the complexity of the ECG attractor is reduced with cardiac muscle ischemia, It may be that D/sub 2/'s of synchronous 12-lead ECG signals may be able to provide valuable diagnostic information about the location of ischemia or infarction. The authors' research shows that the correlation dimension is a valuable parameter in analyzing nonlinear dynamic characteristics of the ECG signal and may provide new insights and methods in diagnosis of cardiac disease.