Spectral analysis and correlation study of vasomotion

Abstract

This thesis consists of two parts focusing on the origin of vasomotion and stochastic resonance in vascular motion, respectively. The first part of the thesis concentrates on the origin of vasomotion. Vasomotion is the spontaneous time-dependent contraction and relaxation of micro arteries with an oscillation frequency of about 0.01 - 0.2 Hz. The physiological mechanisms of vasomotion have not been thoroughly understood. From a dynamics point of view, the heartbeat is the only external loading exerted on the vascular system. Since the period of heartbeat is not constant but variable, we speculated that nonlinear blood flow in the vascular system and variable heartbeat would induce low-frequency vasomotion. We simulated linear flow in a single pipe and nonlinear flow in a network of tubes and found that nonlinear flow would generate low-frequency components. As the vascular system is highly nonlinear, we could deduce that the variable period of heartbeat and the nonlinearity of the vascular system induce vasomotion. Furthermore, using a laser Doppler flowmeter, we measured the time series of radial artery blood flow. Two modified time series were reconstructed with different heartbeat curves but with the same period as the measured time series. Wavelet spectral analysis showed that the low-frequency components were induced by the variable period and independent of the shape of the heartbeat curve. The second part of the thesis focuses on stochastic resonance (SR) in vasomotion. SR is characterized by a system's response to noise and signal-to-noise ratio (SNR). The SNR ascends rapidly and peaks at the point where the SR occurs, after which the SNR descends gradually with increasing noise intensities. We used a laser Doppler flowmeter to measure the time series of radial artery blood flow after adding different disturbances on the forearm. With different cuff pressures added to the radial artery blood flow, the SNR of different frequency bands (~1 Hz, ~0.3 Hz, ~0.1 Hz, 0.03 Hz, and 0.01 Hz) in the pulse signals changed with pressure. We can infer that the ancient Chinese physicians diagnosed by applying pressures on the wrist to sense the SNR changes of different frequency bands in the pulse signals which relates to cardiac, respiratory, myogenic, neurogenic, and endothelial activities. With the disturbance of different noise added to the radial arterial blood flow, the experiment pulse signal was decomposed by HHT. We obtained instantaneous SNR and instantaneous noise from the decomposed IMFs, and the phenomenon of SR was observed in frequency bands around 1 Hz, 0.3 Hz, and 0.1 Hz. The SNR results show that the noise has effects on the blood signal related to the respiratory activities (~0.3 Hz), but has little influence on the signal related to cardiac activities (~1 Hz). Adding white noise and then stopping white noise causes an increase in SNR in the frequency band related to vasomotion (~0.1 Hz).