There are various responses to this question in the literature. It was previously thought that there are one or two main reflection sites (e.g. at the aorto-iliac bifurcation and maybe one in the upper body). However, it is now pretty well established that these are only 'effective' reflection sites. Wave reflection happens wherever there is an impedance mismatch, i.e. due to a local change in vessel stiffness or diameter, or at bifurcations. There is an increasing amount of evidence that wave reflection happens in a gradual or distributed fashion, perhaps due to geometrical or stiffness tapering of the aorta and other vessels, and/or possibly due to the fact that there are many many junctions at different distances from the heart that all contribute a small impedance mismatch. So when doing wave analysis, the best we can do is assess the sum result of many reflections. A couple of papers that might be helpful: Segers (Artery Res, 6(1), 7-11, 2012), Davies (Hypertension, 60(3), 778-785, 2012)

Dear Jonathan Mynard,

Thank you very much for your response. Your recommended papers are good.

I agree that there is no clear physioligical reflection site in the artery system, but the 'effective' reflection sites exist.

As regards to how to assess the wave reflection, there are many methods: 1) wave separation analysis is considered as a gold standard method to assess wave reflection, in which the forward and backward components could be effectively acquired provided that aorta pressure and flow waveforms; 2) augmentation index is widespread used, however, there is still no consensus on the prognostic value of AI in the general population; 3) derivative methods and wave intensity analysis methods are also used to assess the wave reflection; 4) recently, pulse decomposition analysis is used to separtae the pusle into several sub-pulse. We also do some research in this point. (Liu Biomedical Signal Processing and Control, vol. 8, pp. 449-454, 2013).

How do you regard these methods? Which method do you think is promissing for the accurate aeeseement?

Thanks very much.

Each method has its pros and cons. It's always important to remember that any analysis method is based on a model of how we think the arterial system behaves. And every model makes some assumptions and ignores certain phenomena, e.g. often non-linearities, viscoelastic effects etc.

Dear Jonathan Mynard,

Thanks very much. Your answer help me a lot.

It is important to research indeeply at any aspect of this issue and make sure it could contribute for the current studies. I will work on it!