基于多分辨率分析的短时高频心电图近似熵估计

Estimation of Approximate Entropy for Short-term HFECG Based on Multiresolution Analysis

对原始的短时高频心电图(SHFECG)作近似熵计算难以得到稳定、可信的估计值。本研究从小波变换的多分辨率分析出发,对SHFECG信号先作分解,再作近似熵估计,发现含有较高频率成份的心电图细节信号,其近似熵值稳定可靠。并且,对心脏病患者、不同年龄的健康人的SHFECG分析发现不同频率层次的细节信号有着不同的近似熵值分布,显示出该分析方法在心脏疾病的早期诊断中潜在的应用价值。

It is difficult to gain stable and credible values of approximate entropy（ApEn） for original signals of short-term high frequency electrocardiogram（SHFECG）.In this paper,original signals of SHFECG are decomposed with multiresolution wavelet analysis,and then their ApEn is estimated.Results show that stable and credible values of ApEn could be counted from details of these signals.Furthermore,different frequency details of SHFECG have different distributions of ApEn between the patients with heart diseases and healthy objects with a variety of ages,being promising for early diagnosis of heart diseases.