病理心电信号的Jensen-Shannon divergence分析

Complexity analysis of pathological ECG signal based on Jensen-Shannon divergence

目的研究快速区分不同病理心电信号的算法是当前临床医学诊断的研究热点之一。方法本文采用基于Jensen-Shannon divergence的复杂度分析方法,从MIT-BIH标准数据库中提取正常窦性心率、房性早搏、窦性心动过缓信号,分别计算它们的复杂度。结果正常心率信号复杂度最高,窦性心动过缓信号次之,房性早搏最低。方差检验结果表明,此方法得出的3种信号的复杂度具有显著性差异。结论病理心电信号的Jensen-Shannon divergence分析对临床医学检测和诊断房性早搏、窦性心动过缓信号有很好的借鉴意义。

Objective Algorithm research on high-speed diagnosis of pathological ECG is one of the current research hotspots in clinical diagnosis. Methods In this paper, complexity measure based on Jensen- Shannon divergence was used to compute complexity of the ECG signals,which include normal sinus rhythm, atrial premature contraction （APC） and sinus bradycardia （SBR） signals from the MIT-BIH standard database. Results The results showed that the three kinds of signals had different complexity measures. Normal sinus rhythm had the highest complexity ,followed by SBR signals and APC signals. The variance test indicated that this analysis could disclose the significant differences among these three signals＇ complexity. Conclusions This complexity analysis based on Jensen-Shannon divergence is a good reference for clinical detection and diagnosis of APC and SBR signals.