Cardiovascular diseases are the leading cause of mortality worldwide. Central arterial pulse wave velocity (PWV) is a clinical marker of arterial stiffness and an independent predictor of cardiovascular disease and incident hypertension. Local measurements could facilitate widescale clinical use of PWV and improve early cardiovascular diagnostics. However, confluence of forward propagating and reflected arterial waves obscure the spatiotemporal relations of fiducial points on the wave contour, resulting in biased local PWV estimates.
This work introduces the Double Gaussian Propagation Model (DGPM) to measure local PWV in consideration of wave confluence phenomena. The DGPM is a mathematical model that advances Gaussian waveform modeling to full spatiotemporal dimensions, decomposing forward and reflected wave components to obtain unbiased PWV.
For development and evaluation of the DGPM, ultrasound-based distension waveforms from the common carotid artery were acquired from ten human subjects ranging from normotension to hypertension, repeatedly measured at rest and with induced changes in PWV. Per cardiac cycle, the DGPM was fitted to the arterial waveforms by optimizing 8 parameters: amplitude, centroid position, width, and velocity, respectively of the forward and backward propagating wave. Moreover, leveraging on high-quality waveforms, the DGPM was fitted to high-frequency waveform complexes (systolic foot and dicrotic notch) at different pressure levels within the cardiac cycle, rather than the entire pulse contour.
PWV measurements from the DGPM were compared against conventional spatiotemporal PWV (from fiducial point timings and distances), with Bramwell-Hill PWV and continuous noninvasive blood pressure as reference measures. In contrast to conventional spatiotemporal PWV, PWV from the DGPM explains significant variability in the reference Bramwell-Hill PWV, results in lower errors over a wider PWV range, demonstrates significant intra-method consistency, and significantly correlates with blood pressure measures.
In conclusion, the DGPM outperforms conventional spatiotemporal PWV in all investigated respects, potentially enabling practical PWV assessment in routine clinical practice.