Quantitative non-stationary assessment of cardiovascular diseases based on arterial blood pressure waveform by using Hilbert-Huang transform

Analysis of arterial blood pressure (ABP) signal provides valuable information of cardiovascular function. But ABP´s non-stationary characteristics are still unclear and it is hard to evaluate non-stationary characteristics by stationary analysis method, such as Fourier transform and wavelet transform. The aim of the study is to quantitatively evaluate the ABP´s non-stationary characteristics and examine the feasibility of this method by the study of cardiovascular diseases patients. Eight ABP´s intrinsic features were extracted by empirical mode decomposition (EMD) and were evaluated by averaged period and energy density according to the spectral integration method. In the study, 33 subjects (15 healthy subjects, 7 patients with autonomic failure, 11 patients with orthostatic tachycardia syndrome) participated in the passive head-up tilt experiment. The results showed that the characteristics of ABP fluctuation were clearly presented by the changes of IMFs and corresponding averaged periods and energy density. This method was feasible for the qualitative and quantitative assessment of ABP in normal healthy subjects and patients with impaired autonomic regulation of cardiovascular system.