心率变异性的AR模型分析及其在糖尿病患者中的应用

Heart Rate Variability Analysis with AR Model and its Application in Diabetics

针对传统谱估计频率分辨率低和旁瓣泄漏的固有缺点,采用自回归模型功率谱估计对心率变异性信号进行了分析,比较了AR模型方法与FFT方法在心率变异性(HRV)分析中的应用,并对糖尿病患者和正常人的心率变异信号进行了分析和比较。结果表明:自回归模型功率谱估计在心率变异性分析中的有效性优于传统谱估计方法;严重糖尿病患者HRV各频段能量明显减少。

In heart rate variability（HRV）analysis, the classic power spectrum density （PSD） estimate based on FFT is low in frequency resolution and leak in side petal, the PSD estimate based on Auto Regressive （AR） Model is introduced in HRV frequency domain analysis and is compared with PSD based on FFT. The AR model method is used to analyze HRV signals of diabetes and normal. The results show AR model PSD estimate has higher validity than classic PSD estimate; severe diabetes have less HRV power than the normal.