A novel wave reflection model of the human arterial system.

Abstract

A frequency domain distributed 55 segment arterial model was constructed from the reflection perspective to predict pressure waveforms in the large systemic arteries. At any node, the predicted pressure waveform was the combination of a forward propagating waveform and a number of repeatedly reflected waveforms from any possible sites. This approach ensured that any single reflected waveform could be traced back to its origin, and thus the causal-effect relation would be precisely known. This model was evaluated in terms of branch reflection coefficient, terminal vascular bed behavior, and wall viscoelasticity. It was found that the model predicted pressure waveforms were most sensitive to the branch reflection coefficient, and this led to the adoption of the zero-forward reflection assumption at branches. The model-predicted pressure waveforms compared favorably with realistic blood pressure waveforms, especially in the upper limbs. For lower limbs, finer segmentation could further improve the predictions.