对104例冠心病人心电Hoher信号进行心率变异性的分析,计算RR间期序列的近似熵指标的24h分段变化趋势图,并于健康对照组作比较,验证了近似熵这个心率变异性非线性参数的意义。通过LDA(线性鉴别分析)的模式识别方法对病人及健康人的24h变化趋势图进行模式识别和分类,平均正确分类率达99.03%。分类的结果表明,冠心病患者与健康人相比在白天,尤其在早上6点到中午12点间,近似熵指标的降低更明显,利用此时间段作分类正确率更高。
This paper analyzes the approximate entropies of 24-hour Hoher RR data hour by hour. By comparing the trendline of 104 coronary heart disease (CHD) patients anti 54 healthy persons of similar ages,it is proved that the complexity parameter of approximate entropy is very significant. LDA ( linear discriminant analysis) method is utilized to classify the CHD patients and healthy subjects and the average discrimination ratio of this model can reach as high as 99.03%. Results also show that the approximate entropies of CHD patients decrease more obviously than those of healthy persons and the differences in approximate entropies between the two groups becomes more distinguishable during 6 a. m 12 a.m.