IPFM，MIPFM模型产生的仿真心率变异性信号的周期图与AR谱分析研究

为了研究不同心电序列转换方式及不同谱估计方法对心率变异性（ＨＲＶ）信号谱分析结果的影响，本文对积分脉冲频率调制（ＩＰＦＭ）模型及修正积分脉冲频率调制（ＭＩＰＦＭ）模型在输入不同振幅与频率的正弦信号时所产生的随机点过程，用两种心电序列转换方法进行转换得到仿真ＨＲＶ信导；然后，采用周期图与自回归（ＡＲ）谱估计方法计算这种厉真ＨＲＶ信号的功率谱。研究结果表明：①对于ＭＩＰＦＭ模型产生的随机点过程，同一心电序列转换方法所得出的仿真ＨＲＶ信号的ＡＲ谱与周期图的谱峰功率估计基本一致；而对ＩＰＦＭ模型则不完全一致。②ＭＩＰＦＭ模型仿真实验表明，对实际ＨＲＶ信号谱分析，使用低，高频谱峰功率比（ＲＦ）作为反映心脏自主神经张力平衡的指标时，除心电序列传换及谱估计方法可能造成的误差外，当低频谱峰靠近极低频谱峰时，根据ＲＦ值解释生理实验结果会有校大误差。③座分析实际ＨＲＶ信号的工作中，不同心电序列转换方式产生的伪谐波对ＨＲＶ谱分析结果的影响不大。

PERIODOGRAM AND AR SPECTRAL ANALYSISOF SIMULATED HEART RATE VARIABILITYSIGNALS GENERATED BY IPFM ORMIPFM MODELS

Integral pulse frequency modulation( IPFM )model and modifiedIPFM ( MIPFM ) model were used in studies of heart rate variability( HRV ) to generate series of point events When the sinusoidal inputsignals with different amplitudes and frequencies were added on themodels. Simulated HRV signals were obtained through convertion fromseries of point events by use of two types of convertors for electrocar-diogram. Then the power spectral density of simulated HRV signalswas estimated by periodogram and autoregressive spectral estimator,The simulated results showed that: (l ) for the simulated HRV signalsgenerated by the MIPFM model with the same point series convertor,the powers of the main peaks were nearly identical with the periodo-gram and AR spectral estimator,(2)Simulation results based on MIPFMmodel revealed that for the spectral estimation of actual HRV signalswhen the ratio of low frequency power to high frequency power(RF)was used as an index to represent the balance of the cardiac autonomicregulation, the point series convertors and spectral estimators mightcause errors, and more errors would arise if the value of RF was used,to explain the physiological experiniental results when the low frequen-cy peak was close to the much lower frequency peak.( 3) In the ana-lysis of actual HRV signal, the spurious hormonic components generat-ed by different point series convertors produced little effects on thespectral estimation of actual HRV signals.