Study on Pulse-Signal Detection Methods Using Wavelet Transform and Hilbert Huang Transform

The practical wrist pulse of the healthy person and the coronary heart disease patient were analyzed by wavelet transform (WT) and hilbert huang transform ( HHT), and the characteristic information (percentage of energy density) of the pulse signal was discussed. The simulation results show that both WT and HHT are efficient ways in analyzing and processing pulse signal, and can draw main characteristic information from pulse signal. However the basis function is preselected in WT, it doesnt need to be preselected in HHT. In HHT the intrinsic mode function (IMF) is obtained by empirical mode decomposition (EMD), it can reflect the instantaneous frequency of pulse signal, and has the actual physical meaning. The resolving power of time and frequency in WT is restricted by Heisenberg uncertainty principle, and is restricted by each other. While the resolving power of time and frequency in HHT is adaptively changed according to signal intrinsic characteristics. The HHT is more adaptive than WT in analyzing pulse signal. The HHT can offer a new idea to diagnose cardiovascular disease by wrist pulse signal.