HHT方法在脉搏波信号分析中的应用

目的采用HHT(Hibert-HuangTransformation)时间序列分析方法处理从人体采集到的脉搏波信号。方法通过经验模态分解(EMD)技术将一非线性、非稳态过程的原始离散数据序列分解为一组内在模态函数(IMFs),然后对每一个IMF进行HT变换,这样得到的信号幅度和瞬时频率都是时间的函数,即获得脉搏波信号幅度和频率的时间分布。再根据已获得的HH谱,进而得到边际谱。这是一种更具适应性的、新型的、基于模态分解的时间序列数据处理方法。结果首先对一系列由标准的周期函数构造而成的时间序列信号进行了EMD处理,验证HHT方法分解的可行性、有效性;然后分别对一例正常人脉搏波信号和一例典型的冠心病人脉搏波信号进行分解处理,对得到结果进行了比较。结论HHT方法在生物医学信号处理领域将会有广阔的应用前景。

Application of the HHT method to the wrist-pulse-signal analysis

Objective In the paper, the HHT method is adopted to deal with the wrist pulse signals collected from patientbodies. Methods nonlinear and nonstationary series can be decomposed by using the empirical mode decompositionmethod (EMD) into a number of intrinsic mode function (IMF) components. These components are transformed into thoseexpressions called HH spectra which exhibit the amplitude-frequency-time distributions of the data, and then the marginalspectra are obtained by integrating the HH spectra with respect to time. This is a new and applicable time series analysismethod based on mode decomposition. Results A time series constructed from a set of standard cyclic functions isfirst treated with using the EMD method to verify the applicability and effectiveness. And then, two pulse signals, onefrom a healthy person and the other from a patient suffering from the coronary heart disease are decomposed and theirresults are discussed. Conclusion The HHT method has broad prospects in the biomedical engineering and bioinformaticsapplications.