Background: Wave intensity (WI) has traditionally been calculated from measured pressure and velocity waveforms. In the recently proposed reservoir-wave approach, wave intensity is calculated using excess pressure (WI_RW), i.e. the difference between measured and reservoir pressures. Although it has been suggested that WI_RW is more accurate than WI [1], this has not yet been validated in a system with known wave reflection properties.

Methods. This study assessed WI and WI_RW in two one-dimensional models, first, a simple network containing three bifurcations (two producing positive reflection and one producing negative reflection) and second, a full model of the systemic arterial tree. In both, a forward component of pressure was prescribed at the (non-reflecting) inlet, thus forward waves were known a priori. 3-element windkessels formed the model outlets. Modelling results were compared with measurements in the ascending aorta of five adult sheep.

Results. WI accurately predicted the three reflection sites in the simple network, whereas WI_RW did not detect the first or third (positive) sites and overestimated the second (negative) site. In both models, an artefactual mid-systolic forward expansion wave appeared in WI_RW but not WI. In the systemic arterial model, WI_RW predicted predominant negative reflection during systole, whereas WI correctly predicted positive reflection; in vivo results were qualitatively similar.

Conclusion: WI_RW introduces artefactual expansion waves and attenuates or eliminates compression waves. This may limit the utility of WI_RW in the assessment of both forward-running waves and wave reflection in arterial networks.