室性心动过速与室颤的多尺度多重分形分析

Multiscale Multifractality Analysis of Ventricular Tachycardia and Ventricular Fibrillation

基于多尺度多重分形分析(MSMF)对正常心电信号(ECG)、室性心动过速(ventriculartachycard ia,VT)与室颤(ventricular fibrillation,VF)ECG进行研究分析,可以鉴别区分这三种信号。通过直接计算奇异谱f(α)的方法,计算出正常ECG、VT和VF信号的奇异谱,并分析得出其奇异强度分布范围Δα,结果发现这三种信号奇异强度分布范围Δα存在较大差异,并且VF信号的Δα大于VT信号,同时VT信号的Δα又大于正常ECG信号。本文用奇异强度分布范围Δα作为区分正常、VT和VF、ECG的一个判据,且实验证实该判据可有效区分正常ECG、VT和VF信号。

The normal ECG , ventricular tachycardia （VT） and the Ventricular Fibrillation （VF） ECGs are studied based on the Muhiscale Multifractality （MSMF） analysis,which can distinguish between the three kinds of signals. By directly calculating of singularity spectrum f（α） , the areas of the normal ECG, the VT and the VF signals, singularity spectrums were calculated and the singular intensity distribution △α of them were obtained. We found that the singular intensity distribution △α of the VF tended to be larger than that of VT, and the singular intensity distribution △α of the VT tended to be larger than that of normal ECG. Therefore the singular intensity distribution △α could be used as a criterion for identifying normal ECG, VF and VT signals. The experiment demonstrated that the criterion method was effective.