Prony参数模型法在心率变异性谱分析中的应用

PRONY PARAMETRIC MODELING METHOND APPLIED TO SPECTRAL ANALYSIS OF HEART RATE VARIABILITY

本文用衰减正弦函数的线性组合作为人体心率变异性（ＨＲＶ）的信号模型，在不同体位条件下，用Ｐｒｏｎｙ参数模型法估计了ＨＲＶ信号的能量谱，并分析了模型参数的生理意义。结果表明：Ｐｒｏｎｙ，周期图及自回归（ＡＲ）谱在不同体位条件下具有相似的分布特征。立位时，Ｐｒｏｎｙ谱低频谱峰的频率位置向更低频端移动，相应频率衰减因子的绝对值显著减小（Ｐ＜０．０１），振幅有增大趋势；高频谱峰的频率位置基本不变，相应频率衰减因子的绝对值有增大趋势，振幅显著减低（Ｐ＜０．０５）。考虑到体位改变时心血管系统自主神经均衡性的变化，我们认为，本文提出的ＨＲＶ信号模型是合理的，用Ｐｒｏｎｙ方法分析ＨＲＶ信号不仅可以得到ＨＲＶ信号的能量谱，而且可以根据模型参数定量地反映心血管系统自主神经均衡性的变化。

Linear combination of damped sinusoids was used to model the signals of heart rate variability(HRV). The parameters of the damped sinusoids and the energy spectrum were estimated by Prony method in both supine and standing posture. The results showed that the Prony spectrum of HRV signal had similar same distribution characteristics as those of the periodogram and autoregressive(AR) spectrum. On standing, there was a shift of low-frequency(LF) peak of Prony spectrum to the even lower end of frequency abscissa. The absolute value of LF dumping factor was signficantly reduced(P<0.01), and the LF amplitude showed an increased tendency. There was no position change of high-frequency(HF) peak. The absolute value of HF dumping factor showed an increase tendency, and the HF amplitude was significantly decreased (P<0.05).Reference to the facts that the sympathetic tone was increased and the parasympathetic tone was decreased during the posture change from supine to standing, we conclude hat the HRV signal model presented is rational, and by using the prony method, not only the prony spectrum can be obtained, but also cardiovascular autonomic function can be quantitatively assessed.