| Abstract: | Ventricular fibrillation (VF) caused by myo-cardial ischemia is one of the leading factors of death attrib-uted to cardiovascular diseases. It is particularly significantto predict VF and gain valuable time for clinic therapy. Five dogs are taken as the research objects and a VF model is introduced. The nonlinear characteristics of the ECGs before and after VF are investigated with nonlinear multi-parame-ter analysis methods, Gaussian kernel (GK) correlation es-timation algorithm and Lyapunov exponent estimation algo-rithm. Correlation entropy h2 is also presented. The results indicate that there are three parameters which will change at the same time with the conditions of myocardial ischemia,and any changes of a single parameter may be caused byother factors and mislead the judgment. Multi-parameter analysis is more reliable to reveal the heart conditions, and to predict VF without misjudgments. |
