Based On The Matlab Time Frequency Analysis Of Cardiac Electrical Signal Analysis In The Research And Application

Time-frequency analysis mainly includes linear time-frequency analysis and quadratic time-frequency representation. Linear time-frequency analysis can be divided into Short Time Fourier Transform(STFT), Gabor transform and Wavelet transform. And the quadratic time-frequency representation means Wigner-Ville distribution and Cohen time-frequency analysis. The basic idea of STFT transform, as the method of the early time-frequency analysis, is to make window processing of the signal in order to let a short signal be seen as a smooth signal; then analyse the results by adopting Fourier transform. Wigner-Ville distribution, which was proposed in1948, was widely applied in the field of signal analysis, and made up for the shortcomings and deficiencies of the STFT. However, the inevitable cross term, which appeares in the process of application, becomes the major drawback of the Wigner-Ville distribution. The paper makes the Time-frequency analysis of the Electrocardio-signal(ECG) from the following aspects:First, the paper makes the relevant simulated imitation of normal ECG signal and abnormal ECG signals by using the STFT transform and wavelet transform. Through analysing the results of STFT transform, STFT transform has better calculation of the energy distribution of the normal signal and abnormal signal, and clearly finds the difference between the normal signal and abnormal signal from the energy distribution of the signal. However, STFT transform does not show the changing process of energy along with time and frequency. Besides, it can not know the clear time resolution and frequency resolution based on the Heisenberg inequality.Second, the paper calculates the energy distribution of the normal signal and abnormal signal of the ECG signal by adopting the Wigner-Ville distribution, and concludes that ECG signal energy is clearly showed with the frequency changing. It also makes up for the shortcomings of STFT transform. However, Therre is a cross term in the process of analyzing the signals by adopting the Wigner-Ville distribution, which has an effect on the result of analyzing signals. The emphasis of the Wigner-Ville distribution is that eliminating the influences of the signal result caused by the cross term.Third, the kernel fuction of the Wigner-Ville distribution includes the pseudo Wigne-Ville distribution and the smooth pseudo Wigner-Ville distribution. The paper makes the three-dimensinal time-frequency graph and plane time-frequency graph of ECG signal based on thses two kinds of distributions. The results show that the pseudo Wigne-Ville distribution and the smooth pseudo Wigner-Ville distribution can display that the energy of the signal changes with time and frequency, and can eliminate the cross term in a better way, but in the process of eliminating the cross term, the energy of the signal will also be reduced. The pseudo Wigne-Ville distribution and the smooth pseudo Wigner-Ville distribution are the two new time-frequency analysis methods and have a better results of analyzing non-statinary signals.Four, the paper explores the instantaneous frequency ECG signal by using the Hilbert-Huang transform. The result is that the instantaneous frequency of ECG signal mainly changes in the nearby vicinity of QRS, and changes of the normal signal is more regular and abnormal signal changes is irregular. Besides, the paper makes the power spectrum of ECG signal by Burg recursive method. Burg recursive method power spectrum may be a stationary signal and was applied to the non-stationary signals of ECG signal, and finally calculate the change point of frequency.