基于LMCD的病理心电信号分析

目的采用基于Lopez-Mancini-CalbetDivergence（LMCD）的统计复杂度分析方法,对充血性心力衰竭信号、心脏性猝死信号与正常窦性心律信号进行统计复杂度分析。方法采用Bandt-Pompe算法对符号序列进行模式概率统计,分析了充血性心力衰竭信号、心脏性猝死信号与正常窦性心律信号的统计复杂度。结果 3种心律信号的统计复杂度存在差异,正常窦性心率信号的统计复杂度最高,充血性心力衰竭信号次之,心脏性猝死信号最低。方差分析表明,基于LMCD的分析方法得出的3种心电信号的统计复杂度差异具有统计学意义。结论采用LMCD的统计复杂度方法可以有效地区分3种不同生理病理状态下心电信号,为辅助临床诊断提供了一种新手段。

Pathological electrocardiogram analysis based on Lopez-Mancini-Calbet Divergence

Objective To analyze the statistical complexities of congestive heart failure signals,sudden cardiac death signals and normal sinus rhythm signals by the method based on statistical analysis of complexity Lopez-Mancini-Calbet Divergence（LMCD）.Methods The mode probability based on Bandt-Pompe algorithm was analyzed,and the statistical complexity was calculated for congestive heart failure signal,sudden cardiac death signals and normal sinus rhythm signals.Results There were statistical differences in 3 heart rate signals.The complexity of normal sinus rhythm signals was highest,followed by congestive heart failure signals,and that of cardiac sudden death signals was lowest.The variance analysis showed that the significant differences were existed in 3 electrocardiogram（ECG） signal based on LMCD method.Conclusion It is demonstrated that statistical complexity by LMCD could effectively distinguish the ECG in 3 various physiological and pathological states which provide a new means of auxiliary clinical diagnosis.