Pulse Signal Analysis Technique

1 ObjectiveWith the rapid development of pulse diagnosis, more information of Pulse movement and vessel movement researchers can get more difficult it is to find out information which can reflect the function of circulatory system and other physiology systems of subjects. By studying on several popular methods in pulse signals analysis, some implements which give contributions in developing of pulse taking and diagnosis software may come out.2 Methods2.1 A new kind of software which can analyze time domain of pulse signals parameters is developed. Then we evaluated function of it and got time domain of pulse signals parameters by using it. Based on the former researches, we also can identify different kinds of pulse movement by using time domain of pulse signals parameters we got from the software.2.2 We got 39 pulse signals from normal people and 68 pulse signals from patients who were suffered from HPN or CHD. Then every date were cut to 1024 sampling points (sampling frequency is 100Hz) and the oscillation amplitude of every date were normalized (mapping to [0,1]); Secondly, using the EMD method to verify every date(SD=0.2), and got groups of IMF components; Thirdly, calculated the HHS of every date and energy rate beneath nHz (n=1,2,4,5,8,16); At last, calculated the HHS of every IMF components of every date.2.3 We used the WT method to decompose three date came from the formal experiments by"sym8"wavelets base. Then we got d1～d7 as the wavelets coefficients, and reconstructed the signal to 7 components by these wavelets coefficients. After that, we got 7 reconstructed the signal components and we calculated the Hilbert-spectrum of the signal and 7 reconstructed the signal components; At last, we compared the results of this experiment and the results which got by using HHT method.3 Results and Conclusions3.1 The new software can meet the need of analyzing time domain of pulse signals parameters; Time domain of pulse signals parameters analysis method has its advantages that other methods can not shows. It is the basic tool in pulse movement research.3.2.1 HHT method is good at distinguishing the fluent degree of pulse signal and uniformity of pulse rate; It is a good assistant of time domain of pulse signals parameters method.3.2.2 The 2.2 experiment shows that the energy rate(0～8Hz,0～5Hz,0～4Hz,0～2Hz,0～1Hz) of two groups were different by rank-sum test. And the energy rate(0～8Hz,0～5Hz,0～4Hz)of single signal can be used as statistic parameters.3.3.1 WT method is also good at distinguishing the fluent degree of pulse signal; It is a good assistant of time domain of pulse signals parameters method.3.3.2 WT and HHT method both has its advantages and disadvantages. The results come from WT and HHT method are not as affirmative as from time domain of pulse signals parameters analysis method.